DETAILED ACTION
This action is in response to the amendments and remarks filed 5/29/2026. Claims 1-20 are pending and have been examined.
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 6/23/2026 has been entered.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed inventions are directed to non-statutory subject matter without significantly more.
Claim 1
Step 1: The claim recites “A method”, and is therefore directed to the statutory category of process
Step 2A Prong 1: The claim recites the following judicial exception(s)
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This can be performed as a mental process. One can mentally annotate the knowledge graph with relevant statistics from an associated real-time event.
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This can be performed as a mental process. One can merely identify an action performed in the real-time event data, and identify a place in the graph where a corresponding edge should be added.
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This can be performed as a mental process. One can mentally add edges to the knowledge graph corresponding to actions identified in the real-time event data not previously represented.
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: This can be performed as a mental process. One can merely identify streaks and over / underperformance in the data.
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This can be performed as a mental process. One can simply generate and assign a score to each insight by some relevance criteria.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
capturing, by a tracking system of a computing system, play-by-play data for an event: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is insignificant extra-solution activity that precedes the main process of the claim (MPEP 2106.05(g)).
accessing, by the computing system, a database comprising a knowledge graph: This amounts to mere retrieval of data from memory and is insignificant extra-solution activity (MPEP 2106.05(g)).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This amounts to mere data transmission and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
capturing, by a tracking system of a computing system, play-by-play data for an event: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
accessing, by the computing system, a database comprising a knowledge graph: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This is an instance of transmitting data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.).
Claim 2
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites the following further judicial exception(s)
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via templates comprising a deterministic output of descriptive text: This can be performed as a mental process. One can simply devise a set of textual template sentences for interesting types of events, then fill them in by looking through the knowledge graphs.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
generating, by the computing system, the first machine learning model by:
generating a plurality of training data sets based on a plurality of knowledge graphs: This amounts to mere data gathering and is insignificant extra-solution activity (MPEP 2106.05(g)).
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via templates comprising a deterministic output of descriptive text: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
generating, by the computing system, the first machine learning model by:
generating a plurality of training data sets based on a plurality of knowledge graphs: This is an instance of storing information in memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via templates comprising a deterministic output of descriptive text: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
Claim 3
Step 1: The claim recites a process, as in claim 2
Step 2A Prong 1: The claim recites the following further judicial exception(s)
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via the templates comprising the deterministic output of the descriptive text comprises: learning to identify insights that correspond to team-level or play-level streaks: This can be performed as a mental process. One can merely identify consecutive wins by a player or team across multiple graphs.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s).
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via the templates comprising the deterministic output of the descriptive text comprises: learning to identify insights that correspond to team-level or play-level streaks: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s).
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via the templates comprising the deterministic output of the descriptive text comprises: learning to identify insights that correspond to team-level or play-level streaks: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Claim 4
Step 1: The claim recites a process, as in claim 2
Step 2A Prong 1: The claim recites the following further judicial exception(s)
generating, by the computing system by, the second machine learning model by learning, by the second machine learning model, a score for each of the one or more insights by identifying a relevance of each insight compared to other insights: This can be performed as a mental process. One can merely assign a value proportional to the relevance of each insight to some important factor of the event.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
generating, by the computing system by, the second machine learning model by learning, by the second machine learning model, a score for each of the one or more insights by identifying a relevance of each insight compared to other insights: This is mere instruction to apply a judicial exception with a generic computing device (computing system) and a generic data structure (second machine learning model) (MPEP 2106.05(f)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
generating, by the computing system by, the second machine learning model by learning, by the second machine learning model, a score for each of the one or more insights by identifying a relevance of each insight compared to other insights: This is mere instruction to apply a judicial exception with a generic computing device (computing system) and a generic data structure (second machine learning model) (MPEP 2106.05(f)).
Claim 5
Step 1: The claim recites a process, as in claim 4
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a likelihood of occurrence of a particular statistic: This can be performed as a mental process. One can merely assign a value proportional to the likelihood of a statistic related to the insight (e.g. the approximate probability of a pitcher throwing a strike given their past performance).
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s).
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a likelihood of occurrence of a particular statistic: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s).
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a likelihood of occurrence of a particular statistic: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Claim 6
Step 1: The claim recites a process, as in claim 4
Step 2A Prong 1: The claim recites the following further judicial exception(s)
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a particular statistic's impact on a corresponding event: This can be performed as a mental process. One can merely assign a value proportional to the importance of a statistic related to the insight relative to the event’s outcome (e.g. the approximate probability of a team winning if a pitcher throws a strike in the third inning).
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the additional element(s).
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a particular statistic's impact on a corresponding event: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s).
wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a particular statistic's impact on a corresponding event: This is mere instruction to execute a judicial exception with a generic machine learning model (MPEP 2106.05(f)).
Claim 7
Step 1: The claim recites a process, as in claim 1
Step 2A Prong 1: The claim recites no further judicial exception(s)
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the further additional element(s)
wherein presenting, by the computing system, the highest ranking insight of the one or more insights to the one or more end users, comprises interfacing with a client device and prompting the client device to display the highest ranking insight on a display associated therewith: This amounts to mere data transfer and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The further additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
wherein presenting, by the computing system, the highest ranking insight of the one or more insights to the one or more end users, comprises interfacing with a client device and prompting the client device to display the highest ranking insight on a display associated therewith: This is an instance of storing information in memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.)
Claim 8
Step 1: The claim recites “A system”, and is therefore directed to the statutory category of machine
Step 2A Prong 1: The claim recites the following judicial exception(s)
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This can be performed as a mental process. One can mentally annotate the knowledge graph with relevant statistics from an associated real-time event.
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This can be performed as a mental process. One can merely identify an action performed in the real-time event data, and identify a place in the graph where a corresponding edge should be added.
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This can be performed as a mental process. One can mentally add edges to the knowledge graph corresponding to actions identified in the real-time event data not previously represented.
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: This can be performed as a mental process. One can merely identify streaks and over / underperformance in the data.
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This can be performed as a mental process. One can simply generate and assign a score to each insight by some relevance criteria.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
A system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations: This is mere instruction to execute the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
capturing, by a tracking system of a computing system, play-by-play data for an event: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is insignificant extra-solution activity that precedes the main process of the claim (MPEP 2106.05(g)).
accessing, by the computing system, a database comprising a knowledge graph: This amounts to mere retrieval of data from memory and is insignificant extra-solution activity (MPEP 2106.05(g)).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This amounts to mere data transmission and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
A system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations: This is mere instruction to execute the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
capturing, by a tracking system of a computing system, play-by-play data for an event: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
accessing, by the computing system, a database comprising a knowledge graph: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This is an instance of transmitting data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.).
Claims 9-14
Step 1: Claims 9-14 recite a machine as in claim 8.
Step 2A Prong 1: Claims 9-14 recite the same judicial exception(s) as claims 2-7, respectively.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through any additional elements. The analysis of claims 9-14 at this step mirrors that of claims 2-7, respectively, with the exception that claims 9-14 are directed to “A system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations”, said operations mirroring those of claims 2-7. This is a mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s). The analysis of claims 9-14 at this step mirrors that of claims 2-7, with the exception that claims 9-14 are directed to “system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations”, said operations mirroring those of claims 2-7. This is mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Claim 15
Step 1: The claim recites “A non-transitory computer readable medium”, and is therefore directed to the statutory category of article of manufacture
Step 2A Prong 1: The claim recites the following judicial exception(s)
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This can be performed as a mental process. One can mentally annotate the knowledge graph with relevant statistics from an associated real-time event.
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This can be performed as a mental process. One can merely identify an action performed in the real-time event data, and identify a place in the graph where a corresponding edge should be added.
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This can be performed as a mental process. One can mentally add edges to the knowledge graph corresponding to actions identified in the real-time event data not previously represented.
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: This can be performed as a mental process. One can merely identify streaks and over / underperformance in the data.
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This can be performed as a mental process. One can simply generate and assign a score to each insight by some relevance criteria.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through the following additional element(s)
A non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations: This is mere instruction to execute the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
capturing, by a tracking system of a computing system, play-by-play data for an event: This amounts to mere reception of data and is insignificant extra-solution activity (MPEP 2106.05(g)).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is insignificant extra-solution activity that precedes the main process of the claim (MPEP 2106.05(g)).
accessing, by the computing system, a database comprising a knowledge graph: This amounts to mere retrieval of data from memory and is insignificant extra-solution activity (MPEP 2106.05(g)).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This amounts to mere data transmission and is insignificant extra-solution activity (MPEP 2106.05(g)).
Step 2B: The following additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s)
A non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations: This is mere instruction to execute the recited judicial exceptions with generic computer hardware (MPEP 2106.05(f)).
capturing, by a tracking system of a computing system, play-by-play data for an event: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
accessing, by the computing system, a database comprising a knowledge graph: This is an instance of retrieving information from memory, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. iv.).
… wherein the knowledge graph comprises: a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event, and a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event: This merely links the judicial exceptions to a particular field of use (game analysis) (MPEP 2106.05(h)).
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event: This is mere instruction to apply a judicial exception with a generic computing system (MPEP 2106.05(f)).
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges: This is mere instruction to execute a judicial exception with a generic computing machine (MPEP 2106.05(f)).
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges: This is mere instruction to perform a judicial exception with a generic computing machine (MPEP 2105.05(f)).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph, wherein the one or more insights include identifying a streak, an over performing player, an over performing team, an underperforming player, or an underperforming team corresponding to the one or more changes: This is mere instruction to execute a judicial exception by a generic data structure (MPEP 2106.05(f)).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: This is mere instruction to execute a judicial exception with a generic data structure (MPEP 2106.05(f)).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: This is an instance of transmitting data over a network, a limitation known to be well-understood, routine, and conventional (MPEP 2106.05(d) II. i.).
Claims 16-20
Step 1: Claims 16-20 recite an article of manufacture, as in claim 15.
Step 2A Prong 1: Claims 16-20 recite the same judicial exception(s) as claims 2-6, respectively.
Step 2A Prong 2: The judicial exception(s) are not integrated into a practical application through any additional elements. The analysis of claims 16-20 at this step mirrors that of claims 2-6, respectively, with the exception that claims 16-20 are directed to “A non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations”, said operations mirroring those of claims 2-6. This is a mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Step 2B: The additional element(s) of the claim, taken alone or in combination, do not amount to significantly more than the recited judicial exception(s). The analysis of claims 16-20 at this step mirrors that of claims 2-6, with the exception that claims 16-20 are directed to “A non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations”, said operations mirroring those of claims 2-6. This is mere instruction to apply the exceptions using generic computer equipment (MPEP 2106.05(f)).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claim(s) 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Ray (SYSTEM AND METHOD FOR AUTOMATIC GENERATION OF SPORTS MEDIA HIGHLIGHTS, filed 2017, US 2019/0205652 A1), in view of Rabines (SYSTEMS AND METHODS FOR FACILITATING THE GENERATION AND PUBLISHING OF PERSONAL SOCIAL MEDIA, filed 2020, US 2022/0180050 A1).
Regarding claim 1, Ray teaches [a] method, comprising:
capturing, by a tracking system of a computing system, play-by-play data for an event; receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: “The Auto-Highlights Generation Logic may receive inputs from a plurality of data sources, including In-Game AV (audio/visual) media sources (tracking system) (including in-game data and metadata) … The in-game metadata may include: teams played, team home/away information, game date, final score, real time running score with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/ time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play, major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like” (Ray, [0032]). The event in this case is a game. As stated in paragraph [0059] of the instant Specification, play-by-play data can include box score statistics associated with the play. Thus, real time running score is considered a form of real-time play-by-play data.
accessing, by the computing system, a database comprising a knowledge graph: “Referring to FIG. 13, a flow diagram 1300 illustrates one embodiment of a process or logic for implementing graph-based highlight logic, which may be implemented by or part of the In-Game Highlights Logic 42 (FIG. 1). The process/logic begins at block 1302 by retrieving individual activity data graphs (or static graphs) (knowledge graph[s]) for the current game from the Graph Server (database). The data stored on the graph server may be in the format discussed herein for processing by the logic of the present disclosure. For example, the raw static data-graph input data may be in a digital format, equivalent to that shown in the static graphs 502-508 shown in FIG. 5 or 6 with the edges and activity time steps ats … for each of the plays of the game” (Ray, [0088]).
… wherein the knowledge graph comprises:
PNG
media_image1.png
443
333
media_image1.png
Greyscale
(Ray, Figure 5). One static graph from Ray’s system. Each circle (node) represents a player, and each arrow (edge) represents a player action.
a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event:
“The Historical Data Sources 18 provide digital data regarding historical sports event (historical data) and statistics data for the two or more individuals or teams that are playing in the current game (corresponding to the event), which help determine events and patterns of interest for the In-Game Highlights selection rules or weightings used by the In-Game Highlights Logic 42” (Ray, [0034]). The historical data is associated with teams or players in the current game, thus, teams or players in the current game are associated with the historical data.
“Referring to FIG. 5, the four activity snapshots 402-408 shown for the "pick and shoot" basketball play … may be represented by a network data structure (or data network graph or data graph) in FIG. 5 … In that case, the five offensive players (a,b,c,d,e) are shown as solid black circles 510 and the defensive players (defenders) (v,w,x,y,z) are shown as white-center circles 512. Each of the circles 510,512 may be defined as "nodes" (or "vertices") in the data graph … Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data) 516,518, respectively. In this case, the label (or data) format for the nodes 516 is <player location, ball possession>” (Ray, [0046]).
a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event:
“Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data), respectively. In this case, … the label (or data) format for the edges 514 is <player action>, e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc.” (Ray, [0046]). Each edge is associated with an action.
“Referring to FIGS. 9 and 10, the logic of the present disclosure may search the "frequent" static subgraphs of Table 900 to identify "frequent dynamic patterns" or "dynamic normative patterns" (DNPs) within a game. For example, using the Table 900, the logic may compare the frequent static sub-graphs sg1 and sg2, for plays Pl, P2, (or compare the sub-graph graph "edges") to identify common substrings (or common sub-graph edges) for activity time steps ats=1, 2, and 3“ (Ray, [0068]). A dynamic normative pattern (DNP) can be a set of edges common to two subgraphs.
”block 1308 identifies frequent Dynamic Normal Patterns (DNPs) for the current game from the frequent sub-graphs and Play weighting factors (Play Wt) and stores in the DNP Table 1100 (FIG. 11), as discussed herein. Next, a block 1310 identifies current game DNPs closest to DNPs of other games with the same outcome and stores the results in the Graph-Based Highlights Table 1202 (FIG. 12), as discussed herein. For example, referring to FIG.11 and Table 1100, if this game was Game 5 of the 2017 season, the game ended in a Loss for Team A, it was against opponent Team D, which it played before in Game 3 (shown by dashed lines), Team D having a Rank of 6th in the league, and the DNPs identified were DNP7, DNP12, DNP13, and DNP14. The only DNP that appeared in a previous game was DNP7 (historical data corresponding to the event), which also appeared in Game 4 (against Team E) and Game 3 (against Team D). This is particularly relevant because it was against the same opponent (Team D), as discussed herein with FIG. 11.” (Ray, [0094]). Data from a previous game is historical, as it happened prior to the current game, and it’s related to the current event through shared DNPs. A common DNP contains edges present in the previous game, representing previous actions performed in that historical game shared with the current game.
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event:
“The in-game metadata may include: teams played, team home/away information, game date (season statistic), final score, real time running score (up-to-date game statistic) with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play (career statistic), major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like.” (Ray, [0032])
“the raw graphing data logic may process the AV game data and transform it into the data-graph model format discussed herein (e.g., with node and edge labels), which may include applying computer vision to detect the position of players on the court (or field of play), parsing player tracking data to detect position of the player and the ball, and parsing play-by-play data (and metadata) to identify the plays. Any other techniques may be used if desired to obtain the raw static data graph data of the game.” (Ray, [0088])
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges:
“Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic (e.g., block 304, FIG. 3) as part of the In-Game Highlight Logic 42 (FIG. 1) to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed (real-time event data), which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs). To identify Graph-Based Highlights (GBHs) of a game, the game may be viewed as a series of plays, and each play viewed as a series of activities, actions, or movements. Individual activity snapshots (or frames or time steps) of a play or portion thereof (or game strategy), including all (or a predetermined number of) players on the court or field of play, both offensive and defensive entities, may be represented as individual "activity graphs." (Ray, [0043])
“the label (or data) format for the edges 514 is <player action> (first action), e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc. There” (Ray, [0046])
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges:
“As discussed herein, the four "static" data graphs 502-508 (or sg1-sg4, respectively) are shown, for the "pick and shoot" play, where each of the static data graphs 502-508 represents a snapshot in time (ats=1, ats=2, ats=3, ats=4, respectively) during the play, with the nodes and edges having data labels as described herein above, and the corresponding play graph as described hereinafter” (Ray, [0048])
PNG
media_image2.png
792
519
media_image2.png
Greyscale
(Ray, Figure 5); “in the graph 502, the offensive player b (solid black circle with a white ‘b’ in center) 520 has the ball and is located in region 2 of the court, as indicated by the data label [location, ball possession]=[2,1] … defender x located in region 1 without the ball [1,0], … and player b initiates a pass (first action) (as indicated by the label "Pa") on a line 522 to offensive player a, located in region 1 without the ball [1,0]” (Ray, [0048]). In graph 502, an edge representing the pass (‘Pa’) in this time step connects player nodes a and b.
PNG
media_image3.png
799
527
media_image3.png
Greyscale
(Ray, Figure 5); “Next, the graph 504 at activity time step ats=2 shows the defender x moves in region 1, as shown by a line 530, to guard (G) offensive player a, who now has the ball in region 1: [1,0],[1,1], respectively” (Ray, [0049]). In timestep 2, the pass is completed and thus the first action is not represented in the knowledge graph any longer. Player a’s node is updated in response to the ball being passed to them (changing at least one of the plurality of nodes). Additionally, the pass from the previous timestep is completed, so the pass (‘Pa’) edge is removed (changing at least one of the plurality of edges).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic as part of the In-Game Highlight Logic to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed, which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs) (insights)” (Ray, [0043]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights:
“The play weighting factor (Play Wt) (score) is an indication of the significance to the outcome of the play or game (or other significant attribute of the game), and may range from 1 to 5, 1 being lowest significance and 5 being the highest significance, as discussed herein” (Ray, [0092]); “Play Wt (score) may be used to indicate, or be set or adjusted based on, the level of impact, or the likelihood of impact on the game outcome, or on other significant attribute of the game (such as record breaking play, key defensive stop, key strategy, and the like). The value of the Play Wt may be based on multiple elements or effects factored together” (Ray, [0093]). The play weight, an integer valued between 1 and 5 inclusive, may be set based on a significant attribute of the game. No further calculation is required for a single integer.
“Plays may be weighted (or ranked), e.g., Play Wt (score), based on likely importance or significance to the game outcome (or other significant game attribute), as described herein, e.g., a play occurring after the average point-of-no-return time in a game may be given a high weighting (Play Wt) and may thus be given DNP status or otherwise added to the plays selected for the graph-based highlights (GBHs). Also, in some embodiments, the logic may prioritize identification of low ranked teams successful plays and a high ranked teams successful plays, to determine what critical plays or strategies were used to result in the game outcome, such as when there is an unexpected outcome, e.g., high rank team lost to a low ranked team” (Ray, [0075]); “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]). Ray’s system assigns scores to individual plays proportional to how significant they are to the outcome of the game. The determination of whether or not to use a given highlight can be made based on the scores of the play(s) contained within it. The score of a highlight is the sum of the scores of its play(s).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]); “Referring to FIG. 19, a flow diagram 1900 illustrates one embodiment of a process or logic of the present disclosure for providing, among other things, a graphic user interface (GUI) or visualization to the user (end user) on the display of the user/communication device, for receiving and displaying the AV highlight (HLT) content … The process runs when the HLT App is launched and begins at a block 1902, which retrieves data from the Final Highlight Server (or Final HLT Server)” (Ray, [0111]).
Ray relates to automatic sports analytics based on knowledge graphs and is analogous to the claimed invention
While Ray fails to disclose the further limitations of the claim, Rabines teaches a method, comprising:
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “A Storyline (insight) is a newsworthy posting, automatically generated and published on the application. Typically, a Storyline will contain news and report details about an event, such as an upcoming game (i.e., which teams, what date and time), details about the recent actions and performance of real-world protagonists (i.e. specific players' or teams' recent game stats), an editorial assessment regarding the meaning or significance of their recent record or upcoming game (i.e. it may assess a notable outcome in the last matchup ), as well as a premise that defines a line of success ( or failure) by the featured protagonists in a proximal event (i.e. above or below average performance for specific teams or players, in the next scheduled game)“ (Rabines, [0018]); “The server-side application 12 can apply machine learning processes that analyze the data in these tables to find notable patterns of the type that provoke a question or opinion from the traditional NBA fan. For example, the machine learning processes may look for scoring streaks, or whether any player had a "triple double" streak started, which means ten or more points, rebounds and assists. Further, the machine learning processes may apply statistical review to find trends that are other-than-normal, and perhaps even extraordinary … If Mr. Harden will face the Pelicans tonight, the server-side application may identify this upcoming game as a newsworthy event and can use this identified Storyline pattern (insight)” (Rabines, [0070]); “FIG. 5 depicts on example of a publisher processor capable of carrying out the method described above to generate Storylines to aid a user with creating machine displayable content for a data feed published to an account associated with the user … The publishing processor as discussed above has a pattern recognition process (first machine learning model) that will analyze data relevant to those games in the time window and generate … a series of patterns that are candidates for storylines (insights)” (Rabines, [0072]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: “The publishing processor may further include a prioritization processor (second machine learning model) for ranking the candidates identified by the storyline processor into a ranked list of themes … the priority processor can also use those identified relationships to set priority (score) for the candidates for storylines (insights)” (Rabines, [0072]); “In one embodiment, the Storyline machine learning process tracks and analyzes the frequency with which each pattern (insight) appears, to determine which patterns are more or less frequent than others. The Storyline machine learning process provides a higher priority (score) to less frequent patterns (rare patterns). The Storyline machine learning process may also track and analyze the frequency with which users react to certain patterns and give priority to the more popular patterns (storylines that users swipe on the most)” (Rabines, [0060]). The calculated value for priority is inversely proportional to pattern frequency and / or directly proportional to more popular patterns.
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “The systems described herein determine that there is a newsworthy, or comment worthy event (highest ranking) related to topic of shared interest to a large community of users. For example, in the domain of major league sports, the system may determine that a particular NBA basketball player is on a scoring streak of scoring more than 25 points per game. The system may also determine that this pace is statistically exceptional (highest ranking), especially for this player, being, for example, two or three standard deviations above relevant means. The system may process this statistical data into a succinct question, such as will player X's streak of scoring more than 25 points per game continue in tonight's game? … the noteworthy pattern becomes the basis for generation a new posting on the app, while historical data about the topic is used to generate an argument-worthy premise that is included in the posting. Other relevant information is included in the posting and may be published” (Rabines, [0021]).
Rabines relates to automatic sports analytics and is analogous to the claimed invention. Ray teaches a computer system that generates and scores insights based on knowledge graphs. The claimed invention differs from this method by using two machine learning models to generate and score insights, respectively. Rabines teaches a computer system that uses two machine learning models to generate and score insights. Because both Ray and Rabines teach the use of software modules that generate and score insights, it would have been obvious to a person of ordinary skill in the art to substitute Ray’s generation and scoring modules for Rabines’ machine learning generation and scoring modules to achieve the predictable result of automatic generation and scoring of insights using machine learning (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
Regarding claim 2, the rejection of claim 1 is incorporated. The combination of Ray and Rabines further teaches a method, comprising:
generating, by the computer system, the first machine learning model by:
generating a plurality of training data sets based on a plurality of knowledge graphs: Ray: “the present disclosure may use the Graph-Based Highlight Logic … as part of the In-Game Highlight Logic … to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed, which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs) … Individual activity snapshots (or frames or time steps) of a play or portion thereof (or game strategy), including all (or a predetermined number of) players on the court or field of play, both offensive and defensive entities, may be represented as individual ‘activity graphs’” (Ray, [0043]); “once individual static activity graphs have been created (or received or retrieved) for each activity snapshot or time step (a plurality of knowledge graphs), the individual activity graphs (from multiple snapshots) may be combined into a composite or "dynamic" graph (training data set) to represent an entire play (or portion thereof). Each of the plays may also be combined or aggregated to represent an entire game” (Ray, [0052]); “the dynamic graphs (dg's) (plurality of training data sets) for the plays of the game may be analyzed” (Ray, [0054]). Graphs, constructed by the Graph-Based Highlight Logic, are used to learn key insights from a game match as discussed below, hence they’re used to train the model to identify these insights.
learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs: Ray: “the present disclosure may use the Graph-Based Highlight Logic … as part of the In-Game Highlight Logic … to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed, which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs) (insights)” (Ray, [0043]); “the dynamic graphs (dg's) (graphs, each based on a plurality of knowledge graphs) for the plays of the game may be analyzed for "frequent" static sub-graphs, or portions of (or activities in) the dynamic graphs (or series of static graphs) that are repeated during a game at some minimum "frequency" or that are "significant", i.e., are an important or significant part of the play or play outcome, such as, a series of activities that lead to a score, or that involve multiple players with the ball, or that involve the ball, or that occur in a game just before or after a win is assured (e.g., "point of no return"), and the like” (Ray, [0054]). As discussed regarding claim 1, the highlight-generation module taught by Ray can be substituted with the first machine learning model taught by Rabines.
via templates comprising a deterministic output of descriptive text: Rabines: “The publisher processor processes data from the sports data set to identify patterns (insights) within the data set that are associated with a list of predetermined themes … for generating a headline signal representative of a machine displayable string of text and being associated with an identified pattern. The predetermined list of themes may be as described above ‘Hot Streak’, ‘Breakout Rookie’, ‘Highlight Performance’, ‘Big Performance’, ‘Trending Up/Down’; ‘Repeat’, ‘Bounce Back’, ‘Back to Reality’ or any other suitable theme for the sports domain and capable of being associated with a pattern that can be statistically identified from the data set. The publishing processor may further include a prioritization processor (second machine learning model) for ranking the candidates identified by the storyline processor (first machine learning model) into a ranked list of themes” (Rabines, [0072]). The storyline processor and pattern recognition process are equivalent, as evident by their shared number 472 in Rabines’ disclosure. The outputs of the first machine learning model (insights) of Rabines’ system are ranked by the second machine learning model.
Rabines: “The priority processor can send to the editorial headline processor the candidates to be made into Storylines. The editorial headline processor can process the candidate strings, which are typically a string of text, such as ‘Hot Streak’, associated with the identified pattern. The headline processor can alter sections of the string to include a string of event data associated with upcoming event and the patterns, such as by altering the string to ‘Is Harden on a Hot Streak?’” (Rabines, [0072]). The ranked insights are fed into the headline processor, which forms a more detailed string.
Rabines: “In one embodiment, the headline processor stores a series of template strings for each theme, such as the theme ‘Hot Streak’. The template string can include in one example, a string for the theme (descriptive text), such as the text string ‘Will [player X]'s hot streak end [event time]’. The template string can include a formulation of the theme, such as Will X's hot streak end?, as well as replaceable string variables, such as [player X], and [event time]. The headline processor can use a string replacement process to replace the replaceable string variables with text strings that present in a displayable form data (deterministic output) associated with the Storyline, such as the player's name, in this example Mr. Harden, the event time, in this example Wednesday night, or in other examples, it could be tonight at 7 pm?’” (Rabines, [0072]). The detailed string can be formed from a template associated with the theme of the insight, and is used as the final output to the user.
The output of the first machine learning model in Rabines’ system is designed to be fed as input to a template string generator which sends final textual information to a user. This falls within the scope of the first machine learning model learning insights via templates comprising a deterministic output of descriptive text.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the existing combination to use template strings for textual insight output, as disclosed by Rabines. Doing so allows the creation of user-readable descriptive strings that summarize the important aspects of insights, which can be incorporated into more elaborate displayable web forms. See Rabines, [0072], [0074], and Figure 6.
Regarding claim 3, the rejection of claim 2 is incorporated. The combination of Ray and Rabines further teaches a method, wherein learning, by the first machine learning model, the one or more insights based on the plurality of knowledge graphs via the templates comprising the deterministic output of the descriptive text comprises: learning to identify insights that correspond to team-level or play-level streaks:
PNG
media_image4.png
497
1101
media_image4.png
Greyscale
(Rabines, Figure 4);
Rabines: “in step 254 the identified patterns can be checked for whether they include a "Hot Streak?" storyline, wherein such a pattern shows that a protagonist, typically a player, but it can be a coach or other party, has outperformed their statistical average for three games in a row (play-level streak). Such a pattern can be identified using a process such as that depicted in FIG. 4. FIG. 4 depicts pictorially a pattern recognition and learning process 450 (first machine learning model) for identifying a scoring hot streak. FIG. 4 shows a y-axis 451 for points scored, an x-axis for game event and three games, 23, 24, and 25. In the process depicted the learning process (first machine learning model) has learned from monitoring user activity and sports media activity, that scoring streaks over three games during which each game has scoring that is three or more standard deviations 458 from the historical average for the player (historical knowledge), excites interest from the users of the app (insight). Techniques for determining this "hot streak" rule using machine learning are known in the art and may include Generate and Test paradigms that employ historical data (plurality of knowledge) to identify the rule” (Rabines, [0058]). As discussed regarding claim 2, the knowledge graphs are fed into the highlight generation module disclosed by Ray, which can be substituted with the first machine learning model taught by Rabines as discussed regarding claim 1.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the existing combination to identify insights corresponding to streaks, as disclosed by Rabines. Sports reporters pour through volumes of statistics to try and find noteworthy performances, a process automated by Rabines’ application. Hot streaks are exemplary noteworthy performances, considered interesting “plot points” in sports games. Hot streaks also have reliable statistical support and are clearly correct when identified, unlike other insights which may be more complex, nuanced, or harder to gauge. See Rabines, [0027], [0049], and [0060].
Regarding claim 4, the rejection of claim 2 is incorporated. The combination of Ray and Rabines further teaches a method of generating, by the computing system, the second machine learning model by learning, by the second machine learning model, a score for each of the one or more insights by identifying a relevance of each insight compared to other insights: Ray: “Plays may be weighted (or ranked), e.g., Play Wt (score), based on likely importance or significance to the game outcome (or other significant game attribute), as described herein, e.g., a play occurring after the average point-of-no-return time in a game may be given a high weighting (Play Wt) and may thus be given DNP status or otherwise added to the plays selected for the graph-based highlights (GBHs). Also, in some embodiments, the logic may prioritize identification of low ranked teams successful plays and a high ranked teams successful plays, to determine what critical plays or strategies were used to result in the game outcome, such as when there is an unexpected outcome, e.g., high rank team lost to a low ranked team” (Ray, [0075]); “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s])” (Ray, [0108]). As discussed regarding claim 1, the scoring module of Ray can be substituted with the second machine learning model of Rabines.
Regarding claim 5, the rejection of claim 4 is incorporated. The combination of Ray and Rabines further teaches a method, wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a likelihood of occurrence of a particular statistic: Ray: “the dynamic graphs (dg's) for the plays of the game may be analyzed for "frequent" static sub-graphs, or portions of (or activities in) the dynamic graphs (or series of static graphs) that are repeated during a game at some minimum "frequency" or that are "significant" (minimum likelihood of occurrence), i.e., are an important or significant part of the play or play outcome, such as, a series of activities that lead to a score, or that involve multiple players with the ball, or that involve the ball, or that occur in a game just before or after a win is assured (e.g., "point of no return"), and the like. For example, in some embodiments, the present disclosure may give more weight (score) (such as "Play Wt", in Table 900, FIG. 9) to "significant" sub-graphs based on play between average "point of no return" time based on previous games, and the current game's "point of no return" time (discussed more hereinafter).” (Ray, [0054]). As discussed regarding claim 1, the scoring module of Ray can be substituted with the second machine learning model of Rabines.
Regarding claim 6, the rejection of claim 4 is incorporated. The combination of Ray and Rabines further teaches a method, wherein learning, by the second machine learning model, the score for each of the one or more insights by identifying the relevance of each insight compared to other insights comprises: learning to score insights based on a particular statistic's impact on a corresponding event: Ray: “the Plays may be weighted (or ranked) (score[d]), e.g., Play Wt, based on likely importance or significance to the game outcome (or other significant game attribute) (impact on a corresponding event), as described herein” (Ray, [0075]); “The "unexpectedness" of game outcome may be computed by comparing the game outcome with a Bradley-Terry predictive model, which may determine the probability of one team beating another team based on the ability of each team, certain performance factors of each team (e.g., scoring percentage, turnover rate, and the like) and an error correction term” (Ray, [0056]). As discussed regarding claim 1, the scoring module of Ray can be substituted with the second machine learning model of Rabines.
Regarding claim 7, the rejection of claim 1 is incorporated. Ray further teaches a method, wherein presenting, by the computing system, the highest ranking insight of the one or more insights to the one or more end users, comprises: interfacing with a client device and prompting the client device to display the highest ranking insight on a display associated therewith: “on the display screen of the user device (client device) there may be a user settings menu option displayed (not shown), which, when selected, allows the user to select various attributes and features associated with the HLT App software (interfacing), to perform the functions described herein, such as various display options, defaults, and the like, as well as to provide information about the user (user attributes), or for other purposes as described hereinafter” (Ray, [0117]); “The IG-HLT & OGC Content section allows the user to select various aspects of the highlight content … The settings may also set the logic to use only the highest weighted (highest ranking) (or most relevant or significant) components, e.g., OGC Wt or Play Wt. Also, any of the factors, times, windows, settings, weightings (or weighting factors), default values, thresholds, limits, priorities, sequencing, rules, formats, tags, features, sizes, and adjustments discussed herein may also be located in the HLT App settings. Any other IG-HLT & OGC Content settings may be used if desired” (Ray, [0119]).
Regarding claim 8, Ray teaches [a] system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations: “The system, computers, servers, devices and the like described herein have the necessary electronics, computer processing power, interfaces, memory, hardware, software, firmware, logic/state machines, databases, microprocessors, communication links, displays or other visual or audio user interfaces, printing devices, and any other input/ output interfaces, to provide the functions or achieve the results described herein. Except as otherwise explicitly or implicitly indicated herein, process or method steps described herein may be implemented within software modules (or computer programs) executed on one or more general-purpose computers … In addition, a computer-readable storage medium may store thereon instructions that when executed by a machine ( such as a computer) result in performance according to any of the embodiments described herein” (Ray, [0124]).
Said operations comprising:
capturing, by a tracking system of a computing system, play-by-play data for an event; receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: “The Auto-Highlights Generation Logic may receive inputs from a plurality of data sources, including In-Game AV (audio/visual) media sources (tracking system) (including in-game data and metadata) … The in-game metadata may include: teams played, team home/away information, game date, final score, real time running score with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/ time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play, major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like” (Ray, [0032]). The event in this case is a game. As stated in paragraph [0059] of the instant Specification, play-by-play data can include box score statistics associated with the play. Thus, real time running score is considered a form of real-time play-by-play data.
accessing, by the computing system, a database comprising a knowledge graph: “Referring to FIG. 13, a flow diagram 1300 illustrates one embodiment of a process or logic for implementing graph-based highlight logic, which may be implemented by or part of the In-Game Highlights Logic 42 (FIG. 1). The process/logic begins at block 1302 by retrieving individual activity data graphs (or static graphs) (knowledge graph[s]) for the current game from the Graph Server (database). The data stored on the graph server may be in the format discussed herein for processing by the logic of the present disclosure. For example, the raw static data-graph input data may be in a digital format, equivalent to that shown in the static graphs 502-508 shown in FIG. 5 or 6 with the edges and activity time steps ats … for each of the plays of the game” (Ray, [0088]).
… wherein the knowledge graph comprises:
PNG
media_image1.png
443
333
media_image1.png
Greyscale
(Ray, Figure 5). One static graph from Ray’s system. Each circle (node) represents a player, and each arrow (edge) represents a player action.
a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event:
“The Historical Data Sources 18 provide digital data regarding historical sports event (historical data) and statistics data for the two or more individuals or teams that are playing in the current game (corresponding to the event), which help determine events and patterns of interest for the In-Game Highlights selection rules or weightings used by the In-Game Highlights Logic 42” (Ray, [0034]). The historical data is associated with teams or players in the current game, thus, teams or players in the current game are associated with the historical data.
“Referring to FIG. 5, the four activity snapshots 402-408 shown for the "pick and shoot" basketball play … may be represented by a network data structure (or data network graph or data graph) in FIG. 5 … In that case, the five offensive players (a,b,c,d,e) are shown as solid black circles 510 and the defensive players (defenders) (v,w,x,y,z) are shown as white-center circles 512. Each of the circles 510,512 may be defined as "nodes" (or "vertices") in the data graph … Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data) 516,518, respectively. In this case, the label (or data) format for the nodes 516 is <player location, ball possession>” (Ray, [0046]).
a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event:
“Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data), respectively. In this case, … the label (or data) format for the edges 514 is <player action>, e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc.” (Ray, [0046]). Each edge is associated with an action.
“Referring to FIGS. 9 and 10, the logic of the present disclosure may search the "frequent" static subgraphs of Table 900 to identify "frequent dynamic patterns" or "dynamic normative patterns" (DNPs) within a game. For example, using the Table 900, the logic may compare the frequent static sub-graphs sg1 and sg2, for plays Pl, P2, (or compare the sub-graph graph "edges") to identify common substrings (or common sub-graph edges) for activity time steps ats=1, 2, and 3“ (Ray, [0068]). A dynamic normative pattern (DNP) can be a set of edges common to two subgraphs.
”block 1308 identifies frequent Dynamic Normal Patterns (DNPs) for the current game from the frequent sub-graphs and Play weighting factors (Play Wt) and stores in the DNP Table 1100 (FIG. 11), as discussed herein. Next, a block 1310 identifies current game DNPs closest to DNPs of other games with the same outcome and stores the results in the Graph-Based Highlights Table 1202 (FIG. 12), as discussed herein. For example, referring to FIG.11 and Table 1100, if this game was Game 5 of the 2017 season, the game ended in a Loss for Team A, it was against opponent Team D, which it played before in Game 3 (shown by dashed lines), Team D having a Rank of 6th in the league, and the DNPs identified were DNP7, DNP12, DNP13, and DNP14. The only DNP that appeared in a previous game was DNP7 (historical data corresponding to the event), which also appeared in Game 4 (against Team E) and Game 3 (against Team D). This is particularly relevant because it was against the same opponent (Team D), as discussed herein with FIG. 11.” (Ray, [0094]). Data from a previous game is historical, as it happened prior to the current game, and it’s related to the current event through shared DNPs. A common DNP contains edges present in the previous game, representing previous actions performed in that historical game shared with the current game.
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event:
“The in-game metadata may include: teams played, team home/away information, game date (season statistic), final score, real time running score (up-to-date game statistic) with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play (career statistic), major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like.” (Ray, [0032])
“the raw graphing data logic may process the AV game data and transform it into the data-graph model format discussed herein (e.g., with node and edge labels), which may include applying computer vision to detect the position of players on the court (or field of play), parsing player tracking data to detect position of the player and the ball, and parsing play-by-play data (and metadata) to identify the plays. Any other techniques may be used if desired to obtain the raw static data graph data of the game.” (Ray, [0088])
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges:
“Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic (e.g., block 304, FIG. 3) as part of the In-Game Highlight Logic 42 (FIG. 1) to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed (real-time event data), which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs). To identify Graph-Based Highlights (GBHs) of a game, the game may be viewed as a series of plays, and each play viewed as a series of activities, actions, or movements. Individual activity snapshots (or frames or time steps) of a play or portion thereof (or game strategy), including all (or a predetermined number of) players on the court or field of play, both offensive and defensive entities, may be represented as individual "activity graphs." (Ray, [0043])
“the label (or data) format for the edges 514 is <player action> (first action), e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc. There” (Ray, [0046])
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges:
“As discussed herein, the four "static" data graphs 502-508 (or sg1-sg4, respectively) are shown, for the "pick and shoot" play, where each of the static data graphs 502-508 represents a snapshot in time (ats=1, ats=2, ats=3, ats=4, respectively) during the play, with the nodes and edges having data labels as described herein above, and the corresponding play graph as described hereinafter” (Ray, [0048])
PNG
media_image2.png
792
519
media_image2.png
Greyscale
(Ray, Figure 5); “in the graph 502, the offensive player b (solid black circle with a white ‘b’ in center) 520 has the ball and is located in region 2 of the court, as indicated by the data label [location, ball possession]=[2,1] … defender x located in region 1 without the ball [1,0], … and player b initiates a pass (first action) (as indicated by the label "Pa") on a line 522 to offensive player a, located in region 1 without the ball [1,0]” (Ray, [0048]). In graph 502, an edge representing the pass (‘Pa’) in this time step connects player nodes a and b.
PNG
media_image3.png
799
527
media_image3.png
Greyscale
(Ray, Figure 5); “Next, the graph 504 at activity time step ats=2 shows the defender x moves in region 1, as shown by a line 530, to guard (G) offensive player a, who now has the ball in region 1: [1,0],[1,1], respectively” (Ray, [0049]). In timestep 2, the pass is completed and thus the first action is not represented in the knowledge graph any longer. Player a’s node is updated in response to the ball being passed to them (changing at least one of the plurality of nodes). Additionally, the pass from the previous timestep is completed, so the pass (‘Pa’) edge is removed (changing at least one of the plurality of edges).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic as part of the In-Game Highlight Logic to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed, which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs) (insights)” (Ray, [0043]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights:
“The play weighting factor (Play Wt) (score) is an indication of the significance to the outcome of the play or game (or other significant attribute of the game), and may range from 1 to 5, 1 being lowest significance and 5 being the highest significance, as discussed herein” (Ray, [0092]); “Play Wt (score) may be used to indicate, or be set or adjusted based on, the level of impact, or the likelihood of impact on the game outcome, or on other significant attribute of the game (such as record breaking play, key defensive stop, key strategy, and the like). The value of the Play Wt may be based on multiple elements or effects factored together” (Ray, [0093]). The play weight, an integer valued between 1 and 5 inclusive, may be set based on a significant attribute of the game. No further calculation is required for a single integer.
“Plays may be weighted (or ranked), e.g., Play Wt (score), based on likely importance or significance to the game outcome (or other significant game attribute), as described herein, e.g., a play occurring after the average point-of-no-return time in a game may be given a high weighting (Play Wt) and may thus be given DNP status or otherwise added to the plays selected for the graph-based highlights (GBHs). Also, in some embodiments, the logic may prioritize identification of low ranked teams successful plays and a high ranked teams successful plays, to determine what critical plays or strategies were used to result in the game outcome, such as when there is an unexpected outcome, e.g., high rank team lost to a low ranked team” (Ray, [0075]); “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]). Ray’s system assigns scores to individual plays proportional to how significant they are to the outcome of the game. The determination of whether or not to use a given highlight can be made based on the scores of the play(s) contained within it. The score of a highlight is the sum of the scores of its play(s).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]); “Referring to FIG. 19, a flow diagram 1900 illustrates one embodiment of a process or logic of the present disclosure for providing, among other things, a graphic user interface (GUI) or visualization to the user (end user) on the display of the user/communication device, for receiving and displaying the AV highlight (HLT) content … The process runs when the HLT App is launched and begins at a block 1902, which retrieves data from the Final Highlight Server (or Final HLT Server)” (Ray, [0111]).
Ray relates to automatic sports analytics based on knowledge graphs and is analogous to the claimed invention
While Ray fails to disclose the further limitations of the claim, Rabines teaches operations comprising:
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “A Storyline (insight) is a newsworthy posting, automatically generated and published on the application. Typically, a Storyline will contain news and report details about an event, such as an upcoming game (i.e., which teams, what date and time), details about the recent actions and performance of real-world protagonists (i.e. specific players' or teams' recent game stats), an editorial assessment regarding the meaning or significance of their recent record or upcoming game (i.e. it may assess a notable outcome in the last matchup ), as well as a premise that defines a line of success ( or failure) by the featured protagonists in a proximal event (i.e. above or below average performance for specific teams or players, in the next scheduled game)“ (Rabines, [0018]); “The server-side application 12 can apply machine learning processes that analyze the data in these tables to find notable patterns of the type that provoke a question or opinion from the traditional NBA fan. For example, the machine learning processes may look for scoring streaks, or whether any player had a "triple double" streak started, which means ten or more points, rebounds and assists. Further, the machine learning processes may apply statistical review to find trends that are other-than-normal, and perhaps even extraordinary … If Mr. Harden will face the Pelicans tonight, the server-side application may identify this upcoming game as a newsworthy event and can use this identified Storyline pattern (insight)” (Rabines, [0070]); “FIG. 5 depicts on example of a publisher processor capable of carrying out the method described above to generate Storylines to aid a user with creating machine displayable content for a data feed published to an account associated with the user … The publishing processor as discussed above has a pattern recognition process (first machine learning model) that will analyze data relevant to those games in the time window and generate … a series of patterns that are candidates for storylines (insights)” (Rabines, [0072]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: “The publishing processor may further include a prioritization processor (second machine learning model) for ranking the candidates identified by the storyline processor into a ranked list of themes … the priority processor can also use those identified relationships to set priority (score) for the candidates for storylines (insights)” (Rabines, [0072]); “In one embodiment, the Storyline machine learning process tracks and analyzes the frequency with which each pattern (insight) appears, to determine which patterns are more or less frequent than others. The Storyline machine learning process provides a higher priority (score) to less frequent patterns (rare patterns). The Storyline machine learning process may also track and analyze the frequency with which users react to certain patterns and give priority to the more popular patterns (storylines that users swipe on the most)” (Rabines, [0060]). The calculated value for priority is inversely proportional to pattern frequency and / or directly proportional to more popular patterns.
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “The systems described herein determine that there is a newsworthy, or comment worthy event (highest ranking) related to topic of shared interest to a large community of users. For example, in the domain of major league sports, the system may determine that a particular NBA basketball player is on a scoring streak of scoring more than 25 points per game. The system may also determine that this pace is statistically exceptional (highest ranking), especially for this player, being, for example, two or three standard deviations above relevant means. The system may process this statistical data into a succinct question, such as will player X's streak of scoring more than 25 points per game continue in tonight's game? … the noteworthy pattern becomes the basis for generation a new posting on the app, while historical data about the topic is used to generate an argument-worthy premise that is included in the posting. Other relevant information is included in the posting and may be published” (Rabines, [0021]).
Rabines relates to automatic sports analytics and is analogous to the claimed invention. Ray teaches a computer system that generates and scores insights based on knowledge graphs. The claimed invention differs from this method by using two machine learning models to generate and score insights, respectively. Rabines teaches a computer system that uses two machine learning models to generate and score insights. Because both Ray and Rabines teach the use of software modules that generate and score insights, it would have been obvious to a person of ordinary skill in the art to substitute Ray’s generation and scoring modules for Rabines’ machine learning generation and scoring modules to achieve the predictable result of automatic generation and scoring of insights using machine learning (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
Regarding claims 9-14, the rejection of claim 8 is incorporated. As discussed regarding claim 8, Ray teaches [a] system, comprising: a processor; and a memory having programming instructions stored thereon, which, when executed by the processor, causes the system to perform operations. All further limitations of claims 9-14 define operations mirroring those taught by claims 2-7, respectively. A generic machine on its own changes nothing about the method being executed. Thus, claims 9-14 are rejected in view of Ray and Rabines for the same reasons given as for the rejections of claims 2-7, respectively.
Regarding claim 15, Ray teaches [a] non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations: “The system, computers, servers, devices and the like described herein have the necessary electronics, computer processing power, interfaces, memory, hardware, software, firmware, logic/state machines, databases, microprocessors, communication links, displays or other visual or audio user interfaces, printing devices, and any other input/ output interfaces, to provide the functions or achieve the results described herein. Except as otherwise explicitly or implicitly indicated herein, process or method steps described herein may be implemented within software modules (or computer programs) executed on one or more general-purpose computers … In addition, a computer-readable storage medium may store thereon instructions that when executed by a machine (such as a computer) result in performance according to any of the embodiments described herein” (Ray, [0124]).
Said operations comprising:
capturing, by a tracking system of a computing system, play-by-play data for an event; receiving, by the computing system, real-time event data comprising the play-by-play data for the event from the tracking system: “The Auto-Highlights Generation Logic may receive inputs from a plurality of data sources, including In-Game AV (audio/visual) media sources (tracking system) (including in-game data and metadata) … The in-game metadata may include: teams played, team home/away information, game date, final score, real time running score with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/ time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play, major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like” (Ray, [0032]). The event in this case is a game. As stated in paragraph [0059] of the instant Specification, play-by-play data can include box score statistics associated with the play. Thus, real time running score is considered a form of real-time play-by-play data.
accessing, by the computing system, a database comprising a knowledge graph: “Referring to FIG. 13, a flow diagram 1300 illustrates one embodiment of a process or logic for implementing graph-based highlight logic, which may be implemented by or part of the In-Game Highlights Logic 42 (FIG. 1). The process/logic begins at block 1302 by retrieving individual activity data graphs (or static graphs) (knowledge graph[s]) for the current game from the Graph Server (database). The data stored on the graph server may be in the format discussed herein for processing by the logic of the present disclosure. For example, the raw static data-graph input data may be in a digital format, equivalent to that shown in the static graphs 502-508 shown in FIG. 5 or 6 with the edges and activity time steps ats … for each of the plays of the game” (Ray, [0088]).
… wherein the knowledge graph comprises:
PNG
media_image1.png
443
333
media_image1.png
Greyscale
(Ray, Figure 5). One static graph from Ray’s system. Each circle (node) represents a player, and each arrow (edge) represents a player action.
a plurality of nodes, wherein each node of the plurality of nodes represents a player or a team involved in the historical data corresponding to the event:
“The Historical Data Sources 18 provide digital data regarding historical sports event (historical data) and statistics data for the two or more individuals or teams that are playing in the current game (corresponding to the event), which help determine events and patterns of interest for the In-Game Highlights selection rules or weightings used by the In-Game Highlights Logic 42” (Ray, [0034]). The historical data is associated with teams or players in the current game, thus, teams or players in the current game are associated with the historical data.
“Referring to FIG. 5, the four activity snapshots 402-408 shown for the "pick and shoot" basketball play … may be represented by a network data structure (or data network graph or data graph) in FIG. 5 … In that case, the five offensive players (a,b,c,d,e) are shown as solid black circles 510 and the defensive players (defenders) (v,w,x,y,z) are shown as white-center circles 512. Each of the circles 510,512 may be defined as "nodes" (or "vertices") in the data graph … Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data) 516,518, respectively. In this case, the label (or data) format for the nodes 516 is <player location, ball possession>” (Ray, [0046]).
a plurality of edges connecting nodes of the plurality of nodes, wherein each edge of the plurality of edges represents an action performed in historical data corresponding to the event:
“Each of the nodes 510,512 and each of the edges 514 may have an associated label (or data), respectively. In this case, … the label (or data) format for the edges 514 is <player action>, e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc.” (Ray, [0046]). Each edge is associated with an action.
“Referring to FIGS. 9 and 10, the logic of the present disclosure may search the "frequent" static subgraphs of Table 900 to identify "frequent dynamic patterns" or "dynamic normative patterns" (DNPs) within a game. For example, using the Table 900, the logic may compare the frequent static sub-graphs sg1 and sg2, for plays Pl, P2, (or compare the sub-graph graph "edges") to identify common substrings (or common sub-graph edges) for activity time steps ats=1, 2, and 3“ (Ray, [0068]). A dynamic normative pattern (DNP) can be a set of edges common to two subgraphs.
”block 1308 identifies frequent Dynamic Normal Patterns (DNPs) for the current game from the frequent sub-graphs and Play weighting factors (Play Wt) and stores in the DNP Table 1100 (FIG. 11), as discussed herein. Next, a block 1310 identifies current game DNPs closest to DNPs of other games with the same outcome and stores the results in the Graph-Based Highlights Table 1202 (FIG. 12), as discussed herein. For example, referring to FIG.11 and Table 1100, if this game was Game 5 of the 2017 season, the game ended in a Loss for Team A, it was against opponent Team D, which it played before in Game 3 (shown by dashed lines), Team D having a Rank of 6th in the league, and the DNPs identified were DNP7, DNP12, DNP13, and DNP14. The only DNP that appeared in a previous game was DNP7 (historical data corresponding to the event), which also appeared in Game 4 (against Team E) and Game 3 (against Team D). This is particularly relevant because it was against the same opponent (Team D), as discussed herein with FIG. 11.” (Ray, [0094]). Data from a previous game is historical, as it happened prior to the current game, and it’s related to the current event through shared DNPs. A common DNP contains edges present in the previous game, representing previous actions performed in that historical game shared with the current game.
modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event:
“The in-game metadata may include: teams played, team home/away information, game date (season statistic), final score, real time running score (up-to-date game statistic) with time stamps (may also include basic static graph data described herein), play-by-play data (i.e., what play was run at what game clock/time) and special game notable or outcome-related factors such as: overtime (OT), major injury, record breaking play (career statistic), major controversy, fight, technical foul, ejection, fan interference, upset, buzzer beater, sudden-death win/loss, and the like.” (Ray, [0032])
“the raw graphing data logic may process the AV game data and transform it into the data-graph model format discussed herein (e.g., with node and edge labels), which may include applying computer vision to detect the position of players on the court (or field of play), parsing player tracking data to detect position of the player and the ball, and parsing play-by-play data (and metadata) to identify the plays. Any other techniques may be used if desired to obtain the raw static data graph data of the game.” (Ray, [0088])
parsing, by the computing system, the real-time event data to determine a first action corresponding to the plurality of edges:
“Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic (e.g., block 304, FIG. 3) as part of the In-Game Highlight Logic 42 (FIG. 1) to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed (real-time event data), which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs). To identify Graph-Based Highlights (GBHs) of a game, the game may be viewed as a series of plays, and each play viewed as a series of activities, actions, or movements. Individual activity snapshots (or frames or time steps) of a play or portion thereof (or game strategy), including all (or a predetermined number of) players on the court or field of play, both offensive and defensive entities, may be represented as individual "activity graphs." (Ray, [0043])
“the label (or data) format for the edges 514 is <player action> (first action), e.g., pass (Pa), pick (Pi), guard (G), shoot (S), etc. There” (Ray, [0046])
based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges:
“As discussed herein, the four "static" data graphs 502-508 (or sg1-sg4, respectively) are shown, for the "pick and shoot" play, where each of the static data graphs 502-508 represents a snapshot in time (ats=1, ats=2, ats=3, ats=4, respectively) during the play, with the nodes and edges having data labels as described herein above, and the corresponding play graph as described hereinafter” (Ray, [0048])
PNG
media_image2.png
792
519
media_image2.png
Greyscale
(Ray, Figure 5); “in the graph 502, the offensive player b (solid black circle with a white ‘b’ in center) 520 has the ball and is located in region 2 of the court, as indicated by the data label [location, ball possession]=[2,1] … defender x located in region 1 without the ball [1,0], … and player b initiates a pass (first action) (as indicated by the label "Pa") on a line 522 to offensive player a, located in region 1 without the ball [1,0]” (Ray, [0048]). In graph 502, an edge representing the pass (‘Pa’) in this time step connects player nodes a and b.
PNG
media_image3.png
799
527
media_image3.png
Greyscale
(Ray, Figure 5); “Next, the graph 504 at activity time step ats=2 shows the defender x moves in region 1, as shown by a line 530, to guard (G) offensive player a, who now has the ball in region 1: [1,0],[1,1], respectively” (Ray, [0049]). In timestep 2, the pass is completed and thus the first action is not represented in the knowledge graph any longer. Player a’s node is updated in response to the ball being passed to them (changing at least one of the plurality of nodes). Additionally, the pass from the previous timestep is completed, so the pass (‘Pa’) edge is removed (changing at least one of the plurality of edges).
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “Referring to FIGS. 4 and 5, the present disclosure may use the Graph-Based Highlight Logic as part of the In-Game Highlight Logic to perform a graph-based approach to identify critical strategies or high priority events of interest or Highlights or Highlight Events that may contribute to the outcome of the game being analyzed, which may be referred to herein as or Graph-Based Highlights (GBHs or GB-HLTs) (insights)” (Ray, [0043]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights:
“The play weighting factor (Play Wt) (score) is an indication of the significance to the outcome of the play or game (or other significant attribute of the game), and may range from 1 to 5, 1 being lowest significance and 5 being the highest significance, as discussed herein” (Ray, [0092]); “Play Wt (score) may be used to indicate, or be set or adjusted based on, the level of impact, or the likelihood of impact on the game outcome, or on other significant attribute of the game (such as record breaking play, key defensive stop, key strategy, and the like). The value of the Play Wt may be based on multiple elements or effects factored together” (Ray, [0093]). The play weight, an integer valued between 1 and 5 inclusive, may be set based on a significant attribute of the game. No further calculation is required for a single integer.
“Plays may be weighted (or ranked), e.g., Play Wt (score), based on likely importance or significance to the game outcome (or other significant game attribute), as described herein, e.g., a play occurring after the average point-of-no-return time in a game may be given a high weighting (Play Wt) and may thus be given DNP status or otherwise added to the plays selected for the graph-based highlights (GBHs). Also, in some embodiments, the logic may prioritize identification of low ranked teams successful plays and a high ranked teams successful plays, to determine what critical plays or strategies were used to result in the game outcome, such as when there is an unexpected outcome, e.g., high rank team lost to a low ranked team” (Ray, [0075]); “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]). Ray’s system assigns scores to individual plays proportional to how significant they are to the outcome of the game. The determination of whether or not to use a given highlight can be made based on the scores of the play(s) contained within it. The score of a highlight is the sum of the scores of its play(s).
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “a block 1806 creates a Final Highlights Table … and stores the resulting Final Highlights Table in the Final HLT Server. In particular, the block 1806 may select the highlights (H1-Hn) from the HLT Components Table that have In-Game Highlights (IGHs) having plays with the highest weighting factors (Play Wt) (highest ranking insight[s]), or having an Play Wt value greater than a predetermined acceptable Play Wt threshold” (Ray, [0108]); “Referring to FIG. 19, a flow diagram 1900 illustrates one embodiment of a process or logic of the present disclosure for providing, among other things, a graphic user interface (GUI) or visualization to the user (end user) on the display of the user/communication device, for receiving and displaying the AV highlight (HLT) content … The process runs when the HLT App is launched and begins at a block 1902, which retrieves data from the Final Highlight Server (or Final HLT Server)” (Ray, [0111]).
Ray relates to automatic sports analytics based on knowledge graphs and is analogous to the claimed invention
While Ray fails to disclose the further limitations of the claim, Rabines teaches a method, comprising:
generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph: “A Storyline (insight) is a newsworthy posting, automatically generated and published on the application. Typically, a Storyline will contain news and report details about an event, such as an upcoming game (i.e., which teams, what date and time), details about the recent actions and performance of real-world protagonists (i.e. specific players' or teams' recent game stats), an editorial assessment regarding the meaning or significance of their recent record or upcoming game (i.e. it may assess a notable outcome in the last matchup ), as well as a premise that defines a line of success ( or failure) by the featured protagonists in a proximal event (i.e. above or below average performance for specific teams or players, in the next scheduled game)“ (Rabines, [0018]); “The server-side application 12 can apply machine learning processes that analyze the data in these tables to find notable patterns of the type that provoke a question or opinion from the traditional NBA fan. For example, the machine learning processes may look for scoring streaks, or whether any player had a "triple double" streak started, which means ten or more points, rebounds and assists. Further, the machine learning processes may apply statistical review to find trends that are other-than-normal, and perhaps even extraordinary … If Mr. Harden will face the Pelicans tonight, the server-side application may identify this upcoming game as a newsworthy event and can use this identified Storyline pattern (insight)” (Rabines, [0070]); “FIG. 5 depicts on example of a publisher processor capable of carrying out the method described above to generate Storylines to aid a user with creating machine displayable content for a data feed published to an account associated with the user … The publishing processor as discussed above has a pattern recognition process (first machine learning model) that will analyze data relevant to those games in the time window and generate … a series of patterns that are candidates for storylines (insights)” (Rabines, [0072]).
calculating, by the computing system, via a second machine learning model, a score for each of the one or more insights: “The publishing processor may further include a prioritization processor (second machine learning model) for ranking the candidates identified by the storyline processor into a ranked list of themes … the priority processor can also use those identified relationships to set priority (score) for the candidates for storylines (insights)” (Rabines, [0072]); “In one embodiment, the Storyline machine learning process tracks and analyzes the frequency with which each pattern (insight) appears, to determine which patterns are more or less frequent than others. The Storyline machine learning process provides a higher priority (score) to less frequent patterns (rare patterns). The Storyline machine learning process may also track and analyze the frequency with which users react to certain patterns and give priority to the more popular patterns (storylines that users swipe on the most)” (Rabines, [0060]). The calculated value for priority is inversely proportional to pattern frequency and / or directly proportional to more popular patterns.
presenting, by the computing system, a highest ranking insight of the one or more insights to one or more devices of one or more end users: “The systems described herein determine that there is a newsworthy, or comment worthy event (highest ranking) related to topic of shared interest to a large community of users. For example, in the domain of major league sports, the system may determine that a particular NBA basketball player is on a scoring streak of scoring more than 25 points per game. The system may also determine that this pace is statistically exceptional (highest ranking), especially for this player, being, for example, two or three standard deviations above relevant means. The system may process this statistical data into a succinct question, such as will player X's streak of scoring more than 25 points per game continue in tonight's game? … the noteworthy pattern becomes the basis for generation a new posting on the app, while historical data about the topic is used to generate an argument-worthy premise that is included in the posting. Other relevant information is included in the posting and may be published” (Rabines, [0021]).
Rabines relates to automatic sports analytics and is analogous to the claimed invention. Ray teaches a computer system that generates and scores insights based on knowledge graphs. The claimed invention differs from this method by using two machine learning models to generate and score insights, respectively. Rabines teaches a computer system that uses two machine learning models to generate and score insights. Because both Ray and Rabines teach the use of software modules that generate and score insights, it would have been obvious to a person of ordinary skill in the art to substitute Ray’s generation and scoring modules for Rabines’ machine learning generation and scoring modules to achieve the predictable result of automatic generation and scoring of insights using machine learning (MPEP 2143 I. (B) Substituting one known element for another for predictable results).
Regarding claims 16-20, the rejection of claim 15 is incorporated. As discussed regarding claim 15, Ray teaches [a] non-transitory computer readable medium including one or more sequences of instructions that, when executed by one or more processors, causes a computing system to perform operations. All further limitations of claims 16-20 define operations mirroring those taught by claims 2-6, respectively. A generic machine on its own changes nothing about the method being executed. Thus, claims 16-20 are rejected in view of Ray and Rabines for the same reasons given as for the rejections of claims 2-6, respectively.
Response to Arguments
The following responses address arguments and remarks made in the instant remarks dated
05/29/2026.
101 Rejections
On pages 13-15 of the instant remarks, the Applicant argues that the claims are not directed to abstract ideas:
“The pending claims are not directed to an abstract idea because they do not
recite a mental process under Step 2A, Prong One. The Office Action repeatedly
asserts that the claimed limitations, such as capturing play-by-play data, modifying a
knowledge graph, and generating insights via machine learning models, can be
performed mentally. However, this characterization improperly oversimplifies the claims
and ignores their specific technological implementation. The claims require capturing
real-time event data from a tracking system, dynamically updating a knowledge graph
based on that data, and applying a first and second machine learning models to
generate and score insights, a// in real-time. These operations involve continuous
ingestion of streaming data, maintenance and modification of a graph-based data
structure, and execution of trained machine learning models using learned parameters
Under MPEP §2106.04(a)(2), a mental process must be practically performable
in the human mind, which is not the case here. (MPEP § 2106.04(a)(2)(111)(A) "Claims
do not recite a mental process when they do not contain limitations that can practically
be performed in the human mind, for instance when the human mind is not equipped to
perform the claim limitations." Quoting SRI Int'/, Inc. v. Cisco Systems, Inc., 930 F.3d
1295, 1304 (Fed. Cir. 2019). The USPTO's August 4, 2025, memorandum on subject
matter eligibility1 further clarified that a claim does not recite a mental process where the
claimed operations cannot be practically performed in the human mind. Machine
learning models, including trained models applied to large-scale and real-time datasets,
inherently require computational processes that are not practically performable
mentally. Machine learning models employ complex algorithms to process large-scale
datasets derived from real-time video streams, enabling continuous analysis and
extraction of meaningful patterns at speeds and scales that cannot be achieved by
human cognition. Accordingly, the Examiner's characterization of the claimed machine
learning limitations as mental processes is inconsistent with USPTO guidance, which
requires that claims be evaluated as a whole and prohibits oversimplifying technically
specific operations into generalized mental steps.”
In regards to the Applicant’s arguments above, the Examiner respectfully disagrees that the claims, as amended, recite no mental processes. As stated in MPEP 2106.04(a)(2)(III), The courts do not distinguish between mental processes that are performed entirely in the human mind and mental processes that require a human to use a physical aid (e.g., pen and paper or a slide rule) to perform the claim limitation. See, e.g., Benson, 409 U.S. at 67, 65, 175 USPQ at 674-75, 674 … Nor do the courts distinguish between claims that recite mental processes performed by humans and claims that recite mental processes performed on a computer. As the Federal Circuit has explained, "[c]ourts have examined claims that required the use of a computer and still found that the underlying, patent-ineligible invention could be performed via pen and paper or in a person’s mind." Versata Dev. Group v. SAP Am., Inc., 793 F.3d 1306, 1335, 115 USPQ2d 1681, 1702 (Fed. Cir. 2015). See also Intellectual Ventures I LLC v. Symantec Corp., 838 F.3d 1307, 1318, 120 USPQ2d 1353, 1360 (Fed. Cir. 2016) (‘‘[W]ith the exception of generic computer-implemented steps, there is nothing in the claims themselves that foreclose them from being performed by a human, mentally or with pen and paper.’’); Mortgage Grader, Inc. v. First Choice Loan Servs. Inc., 811 F.3d 1314, 1324, 117 USPQ2d 1693, 1699 (Fed. Cir. 2016) (holding that computer- implemented method for "anonymous loan shopping" was an abstract idea because it could be "performed by humans without a computer").
Claim 1 recites limitations amounting to mental processes performed on generic computer components and generic data structures tantamount to generic computer components, both insufficient to render a mentally performable task non-abstract. For example, claim 1 recites the limitation “generating, by the computing system, via a first machine learning model, one or more insights based on the modified knowledge graph”, reciting a mental process of generating insights based on a graph performed at a high level by a machine learning model, a generic data structure insufficient to render the limitation non-abstract.
Similar reasoning is applicable to the rest of the claims. See the 101 rejections section for more detail. The Examiner asserts that the claims, as amended, recite mental processes, and maintains their rejections on these grounds.
On pages 15-18 of the instant remarks, the Applicant argues that the recited judicial exceptions are practically integrated through improvements to existing technology:
“The claim recites additional elements that integrate any alleged judicial exception
into a practical application by reciting an improvement in the functioning of a computer
or an improvement to other technology or technical field. See MPEP § 2106.04(d)(1)
("[o]ne way to demonstrate such integration is when the claimed invention improves the
functioning of a computer or improves another technology or technical field. The
application or use of the judicial exception in this manner meaningfully limits the claim
by going beyond generally linking the use of the judicial exception to a particular
technological environment, and thus transforms a claim into patent-eligible subject
matter.") The claim elements recite improvements to the technical field of automatically
generating event insights during a live event, thereby improving upon conventional
static rule-based approaches and producing better quality and in-depth insights. (Spec.
at[0018].)
…
The Examiner has alleged that the claimed invention constitutes mere
automation of manual processes, citing MPEP 2106.05(a)(I). Applicant respectfully
disagrees. The claims do not merely automate a manual process; rather, they recite a
specific technical architecture comprising: (a) a knowledge graph data structure that is
dynamically modified with up-to-date game statistics, season statistics, and career
statistics based on real-time streaming data, as recited in claim 1; (b) a first machine
learning model that generates insights from the modified knowledge graph; and (c) a
second machine learning model that scores those insights. This ordered combination
represents a specific technical implementation, not mere automation of what a human
analyst could do. The Examiner has also alleged that the claims do not reflect
improvements over static rule-based approaches. Applicant respectfully disagrees. The
specific claim limitations, namely, the dynamic knowledge graph modification step that
updates the knowledge graph with up-to-date game statistics, season statistics, and
career statistics, combined with the dual machine learning model architecture, are the
very limitations that distinguish the claims from static rule-based approaches. These
limitations enable the system to identify insights that were not previously identified by
human analysts or static rules, precisely because the knowledge graph is continuously
updated with cumulative statistical data and the machine learning models can identify
patterns across this dynamically updated data structure that would not be apparent from
static rules or human analysis alone”
In response to the Applicant’s argument that the recited judicial exceptions of the claimed invention are practically integrated through an improvement to existing technology or technical field, the Examiner notes that improvements cannot be made through a recited judicial exception. As noted by MPEP 2106.05(a), It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. See the discussion of Diamond v. Diehr, 450 U.S. 175, 187 and 191-92, 209 USPQ 1, 10 (1981)) in subsection II, below. In addition, the improvement can be provided by the additional element(s) in combination with the recited judicial exception. See MPEP § 2106.04(d) (discussing Finjan, Inc. v. Blue Coat Sys., Inc., 879 F.3d 1299, 1303-04, 125 USPQ2d 1282, 1285-87 (Fed. Cir. 2018)).
The Applicant is arguing improvement through the use of a knowledge graph from which insights can be derived. However, as noted in the rejection for claim 1 under 35 U.S.C. 101, modifying the knowledge graph with statistics, modifying the graph based on actions not being performed, generating insights based on the modified graph, and scoring said insights can all be performed as mental processes. The improvements argued by the Applicant come from these mental processes, not the additional elements of the claim. While the claimed invention contains additional elements, they are insufficient to provide the argued improvements to existing technology or technical fields. For example, and as noted in more detail in the 101 rejections section, executing these mental processes with a generic computer system or a generic machine learning models amounts to mere instruction to execute judicial exceptions with generic computer components, and accessing event data or the graph itself amounts to mere data retrieval and is insignificant extra-solution activity.
Thus, no rejections are withdrawn on these grounds.
103 Rejections
On pages 18-19 of the instant remarks, the Applicant argues that the relied upon references fail to fully disclose the amended claims:
“The cited references fail to disclose or suggest, and the rejection fails to otherwise consider, each and every element of the rejected claims. For example, the cited references fail to disclose or
suggest, inter alia, "modifying, by the computing system, the knowledge graph with up-to-
date game statistics, season statistics, and career statistics for entities involved in the
real-time event" and "based on a determination that the first action is not
represented in the knowledge graph, modifying ... the knowledge graph,
wherein the modifying includes changing at least one of the plurality of nodes or the at least one of
the plurality of edges" as currently recited in independent claim 1.
…
Moreover, Ray's graphs track spatial player positions and individual player
actions (pass, pick, guard, shoot) within individual plays-they are play-level activity
snapshots showing where players are located on the court and what actions they
perform during a single play sequence. By contrast, claim 1 as amended recites
"modifying, by the computing system, the knowledge graph with up-to-date game
statistics, season statistics, and career statistics for entities involved in the real-time
event." This is a fundamentally different concept: maintaining cumulative statistical data
across games, seasons, and careers within the knowledge graph, not merely recording
spatial positions during a single play. Ray's static graphs do not maintain or update
game statistics, season statistics, or career statistics-they simply capture player
positions and actions at discrete moments within a play. Rabines discloses a pattern
recognition process that analyzes tabular sports data to identify storyline patterns, but
does not teach or suggest modifying a knowledge graph with up-to-date game statistics,
season statistics, and career statistics for entities involved in a real-time event.
Accordingly, Rabines does not cure this deficiency.
Rabines fails to cure Ray's deficiencies, and therefore claim 1 cannot be
rendered obvious in view of Ray and Rabines either individually or in combination.
Coiner was relied upon for limitations that have been removed from the amended claims
and is therefore moot.”
Regarding the Applicant’s arguments above, the Examiner respectfully disagrees.
Regarding “modifying, by the computing system, the knowledge graph with up-to-date game statistics, season statistics, and career statistics for entities involved in the real-time event”, as recited in amended claim 1, Ray discloses generating static graphs using play-by-play data and metadata (Ray, [0088]), the metadata potentially including game dates (season statistics), real-time running scores (up-to-date statistics) of games, and record breaking plays (career statistics) (Ray, [0032]).
Regarding “based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges”, Ray discloses removing action edges from subsequent static graphs (time steps in a dynamic graph, equivalently) after the action has been completed and is no longer represented, and modifying nodes accordingly (Ray, [0048-0049], Figure 5).
All of the amended claims are found to be obvious over Ray in view of Rabines. See the 103 rejections section for more detail. No rejections are withdrawn on these grounds.
On page 19 of the instant remarks, the Applicant argues that Ray fails to disclose modifying a knowledge graph:
“The Office Action alleges that Ray teaches modifying a knowledge graph. (Office
Action at pp. 33-36.) Ray does not teach modifying a knowledge graph. Ray teaches
"four 'static' data graphs 502-508 (or sg1-sg4, respectively) ... for the "pick and shoot"
play, where each of the static data graphs 502-508 represents a snapshot in time
(ats=1, ats=2, ats=3, ats=4, respectively) during the play." (Ray at [0048].) Each data
graph is a single "snapshot" in time of the 'pick and shoot' play. Id. Ray does not teach
changing one or any of the four static data graphs. Nothing is modified, hence the term
"static" to describe the four data graphs. Therefore, Ray does not teach or suggest
modifying a knowledge graph as recited in the claims, nor does Ray teach modifying the
knowledge graph "based on a determination that the first action is not represented" as
recited in the claim. Because Ray's graphs are static and illustrate four progressive
moments of the pick and shoot play, a decision to modify one or more of the four data
graphs is not and would not be taught or described by Ray.”
Regarding the assertion that Ray’s static graph generation is not commensurate with modification of a knowledge graph, as claimed, the Examiner respectfully disagrees. While Ray does disclose a series of static activity graphs, these individual graphs may be seen as different stages of one ‘dynamic graph’ at different time steps, which Ray makes clear in paragraph [0052]. With that in mind, each static graph is representative of the larger ‘dynamic graph’ at a particular time step, and the static graphs together comprise an ordered sequence of modifications for said dynamic graph. The Examiner sees no substantial difference between the graphs disclosed by Ray and those of the claimed invention.
Regarding “based on a determination that the first action is not represented in the knowledge graph, modifying, by the computing system, the knowledge graph, wherein the modifying includes changing at least one of the plurality of nodes or the at least one of the plurality of edges”, this limitation is disclosed in its entirety by Ray, as discussed in previous responses above.
No rejections are withdrawn on these grounds. See the 103 rejections section for more detail.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Huang (Graph Analysis of Major League Soccer Networks: CS224W Final Project, 2018, Stanford University) teaches the construction of a knowledge graph with nodes representing players in a game of sports and edges representing actions between them
Grandin et al. (Signal Capture And Distribution System, published 4/23/2002, US 6378132 B1) discloses a method of annotating play-by-play information from sporting events
Gleadall (Automated statistics content preparation, filed 2012, US20140142921A1) teaches a method of automatically identifying, ranking, and presenting sporting events
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Aaron P Gormley whose telephone number is (571)272-1372. The examiner can normally be reached Monday - Friday 12:00 PM - 8:00 PM EST.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Michelle T Bechtold can be reached at (571) 431-0762. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/AG/Examiner, Art Unit 2148 /MICHELLE T BECHTOLD/ Supervisory Patent Examiner, Art Unit 2148
Prosecution Insights