DETAILED ACTION
Status of the Application
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA . This communication is a final action in response to communications filed on 1/2/2026. Claims 1-20 are currently pending. Claims 11-20 are withdrawn in response to a restriction requirement. Claims 1-10 have been considered below.
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 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. Claims 1-10 are determined to be directed to an abstract idea.
The claims 1-10 are directed to a judicial exception (i.e., law of nature, natural phenomenon, or abstract idea), without providing a practical application integration and without providing significantly more.
As per Step 1 of the subject matter eligibility analysis, Claims 1-10 are directed to an apparatus (i.e., machine), which is one of the statutory categories of invention.
As per Step 2A-Prong 1 of the subject matter eligibility analysis, Claim 1 recites specifically to the abstract idea of feature generation for entities by identifying a dataset corresponding to an entity, the dataset comprising a plurality of characteristics of the entity, a plurality of historic actions performed by the entity, and a plurality of historic events involving the entity, the plurality of historic actions associated with one or more products or services; generating a hierarchy data structure corresponding to the entity based on the dataset corresponding to the entity, the hierarchy data structure defining a first data level for a first subset of the dataset and a second data level for a second subset of the dataset; generating a feature set based on a first featurization process applied to the first subset and a second featurization process applied to the second subset combining outputs of the first and second featurization processes into a combined feature set; training, by the one or more processors, a model associated with the entity using the combined feature set as training data, wherein the combined feature set is generated from different featurization processes applied to different hierarchy levels of entity data; and executing a model associated with the entity using the feature set as input, all of which include mental processes (observing and evaluating data regarding entities to make a judgement/opinion on generating feature sets), and certain methods of organizing human activities, specifically fundamental economic practice (generating feature sets from entity/business data possibly for managing future acts of the entity better), managing personal behavior and interactions (following rules and instructions to generate and utilize feature sets from data associated to entities). Claims 2-10 are further directed to the abstract idea of performing functions provided above for claim 1 with further details/embellishments of the abstract idea. After considering all claim elements, both individually and in combination and in ordered combination, it has been determined that the claims do not amount to significantly more than the abstract idea itself.
As per Step 2A-Prong 2 of the subject matter eligibility analysis, while the claims 1-10 recite additional limitations which are hardware or software elements, such as automated {feature generation}, one or more processors coupled to a non-transitory memory, machine-learning, these limitations are not enough to qualify as a practical application being recited in the claims along with the abstract idea since these elements are merely invoked as a tool to apply instructions of an abstract idea in a particular technological environment, and mere application of an abstract idea in a particular technological environment and merely limiting the use of an abstract idea to a particular technological field do not integrate an abstract idea into a practical application (MPEP 2106.05(f)&(h)). The claims do not amount to "practical application" for the abstract idea because they neither (1) recite any improvements to another technology or technical field; (2) recite any improvements to the functioning of the computer itself; (3) apply the judicial exception with, or by use of, a particular machine; (4) effect a transformation or reduction of a particular article to a different state or thing; (5) provide other meaningful limitations beyond generally linking the use of the judicial exception to a particular technological environment.
As per Step 2B of the subject matter eligibility analysis, while the claims 1-10 recite additional limitations which are hardware or software elements, such as automated {feature generation}, one or more processors coupled to a non-transitory memory, machine-learning, these limitations are not enough to qualify as “significantly more” being recited in the claims along with the abstract idea since these elements are merely invoked as a tool to apply instructions of an abstract idea in a particular technological environment, and mere application of an abstract idea in a particular technological environment and merely limiting the use of an abstract idea to a particular technological field do provide significantly more to an abstract idea (MPEP 2106.05 (f) & (h)). The claims do not amount to "significantly more" than the abstract idea because they neither (1) recite any improvements to another technology or technical field; (2) recite any improvements to the functioning of the computer itself; (3) apply the judicial exception with, or by use of, a particular machine; (4) effect a transformation or reduction of a particular article to a different state or thing; (5) add a specific limitation other than what is well-understood, routine and conventional in the field; (6) add unconventional steps that confine the claim to a particular useful application; nor (7) provide other meaningful limitations beyond generally linking the use of the judicial exception to a particular technological environment.
Therefore, since there are no limitations in the claims 1-10 that transform the exception into a patent eligible application such that the claims amount to significantly more than the exception itself, and looking at the limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually, the claims are rejected under 35 USC § 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 102
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1-10 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Simmons et al (US 20110040613 A1).
As per Claim 1, Simmons teaches a method of automated feature generation for entities (para. 0005) comprising:
identifying, by one or more processors coupled to a non-transitory memory, a dataset corresponding to an entity, the dataset comprising a plurality of characteristics of the entity, a plurality of historic actions performed by the entity, and a plurality of historic events involving the entity, the plurality of historic actions associated with one or more products or services (para. 0164, 0174, “Further, in embodiments of the invention, the learning machine facility 138 may utilize various types of algorithms to refine the economic valuation models of the real-time bidding machine facility 142. The algorithms may include, without any limitations, decision tree learning, association rule learning, artificial neural networks, genetic programming, inductive logic programming, support vector machines, clustering, Bayesian networks, and reinforcement learning. ”; para. 0150, “Further, in embodiments of the invention, a data integration facility 134 and the contextualizer service facility 132 may be associated with the analytics platform facility 114 and the one or more databases. The data integration facility 134 may facilitate the integration of different types of data from one or more databases into the analytics platform facility 114. The contextualizer service facility 132 may identify the contextual category of a medium for advertising and/or publisher content, website, or other publisher ad context. In an example, a contextualizer may analyze web content to determine whether a web page contains content about sports, finance, or some other topic. This information may be used as an input to the learning machine facility in order to identify the relevant publishers and/or web pages where ads may appear. In another embodiment, the location of the ad on the publisher 112 web page may be determined based on the information. In an embodiment of the invention, the contextualizer service facility 132 may also be associated with the real-time bidding machine facility 142 and/or with the one or more databases.”; para. 0005-0006, “The machine learning may be based at least in part on a third party dataset… the third party dataset may include data relating to users of advertising content, contextual data relating to the plurality of available placements, and financial data relating to historical advertisement impressions. The contextual data may derive from a contextualizer service that may be associated with the machine learning facility. In embodiments, data relating to users of advertising content may include demographic data, transaction data, and advertisement conversion data. In embodiments, the economic valuation model may be based in part on real time event data, historic event data, user data, third-party commercial data, advertiser data, and advertising agency data”; also see para. 0055-0056, regarding the databases providing add related data and historical event data to the machine learning; para. 0203-0205 regarding the system structure of the invention including processors and memories);
generating, by the one or more processors, a hierarchy data structure corresponding to the entity based on the dataset corresponding to the entity, the hierarchy data structure defining a first data level for a first subset of the dataset and a second data level for a second subset of the dataset (para. 0170, “in an embodiment of the invention, the real-time bidding machine facility 142 may use a primary model for predicting an economic valuation of each of a plurality of available web publishable advertisement placements based in part on past performance and prices of similar advertisement placements. The real-time bidding machine facility 142 may also use a second model for predicting an economic valuation of each of the plurality of web publishable advertisement placements. After predicting the economic valuations using both the primary model and the second model, the real-time bidding machine facility 142 may compare the valuations produced by the primary model and the second model to determine a preference between the primary model and the second model. In an embodiment of the invention, the comparison of the valuations may include retrospectively comparing the extent to which the models reflect actual economic performance of advertisements. Further, in an embodiment of the invention, the primary model may be an active model responding to purchase requests. The purchase request may be a time limited purchase request. In an embodiment of the invention, the second model may replace the primary model as the active model responding to purchase requests. Further, the replacement may be based on a prediction that the second model may perform better than the primary model under the current market conditions. In embodiments of the invention, the prediction may be based at least in parts on machine learning, historical advertising performance data 130, historical event data, and real-time event data 160.”; para. 0203-0205 regarding the system structure of the invention including processors and memories);
generating, by the one or more processors, a feature set based on a first featurization process applied to the first subset and a second featurization process applied to the second subset combining outputs of the first and second featurization processes into a combined feature set (para. 0170, “in an embodiment of the invention, the real-time bidding machine facility 142 may use a primary model for predicting an economic valuation of each of a plurality of available web publishable advertisement placements based in part on past performance and prices of similar advertisement placements. The real-time bidding machine facility 142 may also use a second model for predicting an economic valuation of each of the plurality of web publishable advertisement placements. After predicting the economic valuations using both the primary model and the second model, the real-time bidding machine facility 142 may compare the valuations produced by the primary model and the second model to determine a preference between the primary model and the second model. In an embodiment of the invention, the comparison of the valuations may include retrospectively comparing the extent to which the models reflect actual economic performance of advertisements. Further, in an embodiment of the invention, the primary model may be an active model responding to purchase requests. The purchase request may be a time limited purchase request. In an embodiment of the invention, the second model may replace the primary model as the active model responding to purchase requests. Further, the replacement may be based on a prediction that the second model may perform better than the primary model under the current market conditions. In embodiments of the invention, the prediction may be based at least in parts on machine learning, historical advertising performance data 130, historical event data, and real-time event data 160.”; para. 0137, “In embodiments, the learning system may create rules and algorithms, as described herein, using training data sets, such as that derived from a prior retailer advertising campaign. The training dataset may include a record of prior impressions, conversions, actions, clickthroughs and the like performed by a plurality of digital media users with the advertisements that were included in the prior campaign.”; para. 0203-0205 regarding the system structure of the invention including processors and memories); and
training, by the one or more processors, a machine-learning model associated with the entity using the combined feature set as training data, wherein the combined feature set is generated from different featurization processes applied to different hierarchy levels of entity data (para. 0137, “In embodiments, the learning system may create rules and algorithms, as described herein, using training data sets, such as that derived from a prior retailer advertising campaign. The training dataset may include a record of prior impressions, conversions, actions, clickthroughs and the like performed by a plurality of digital media users with the advertisements that were included in the prior campaign.”; para. 0164-0195, regarding the learning machine facility refining evaluation models, replacing/prioritizing models based on past performance on first and second ad campaigns and their datasets; para. 0073-0075, 0118, 0127, 0137, regarding algorithm training; para. 0139, “The opportunistic updating component may select what is the order and priority for replacing sections of the valuation model. Such prioritization may be based on the economic valuation of the section to replace versus the new section to incorporate. As a result the system may create a prioritized set of instructions as to what data or sections of the model to add to the valuation system and in what order to do so.” Wherein the prioritization is equivalent to hierarchy levels; para. 0203-0205 regarding the system structure of the invention including processors and memories);
executing, by the one or more processors, a machine-learning model associated with the entity using the feature set as input (para. 0164-0195, regarding the learning machine facility refining evaluation models, replacing/prioritizing models based on past performance on first and second ad campaigns and their datasets; para. 0203-0205 regarding the system structure of the invention including processors and memories).
As per Claim 2, Simmons teaches a method as provided in claim 1 above. Simmons further teaches generating, by the one or more processors, an explanation value based on an output of the machine-learning model (para. 0164-0195, regarding the learning machine facility refining evaluation models, replacing/prioritizing models based on past performance on first and second ad campaigns and their datasets).
As per Claim 3, Simmons teaches a method as provided in claim 1 above. Simmons further teaches wherein the plurality of characteristics further comprise at least one second characteristic of a second entity associated with the entity, and the plurality of historic actions comprise at least one second historic action performed by the second entity (para. 0170, “in an embodiment of the invention, the real-time bidding machine facility 142 may use a primary model for predicting an economic valuation of each of a plurality of available web publishable advertisement placements based in part on past performance and prices of similar advertisement placements. The real-time bidding machine facility 142 may also use a second model for predicting an economic valuation of each of the plurality of web publishable advertisement placements. After predicting the economic valuations using both the primary model and the second model, the real-time bidding machine facility 142 may compare the valuations produced by the primary model and the second model to determine a preference between the primary model and the second model. In an embodiment of the invention, the comparison of the valuations may include retrospectively comparing the extent to which the models reflect actual economic performance of advertisements. Further, in an embodiment of the invention, the primary model may be an active model responding to purchase requests. The purchase request may be a time limited purchase request. In an embodiment of the invention, the second model may replace the primary model as the active model responding to purchase requests. Further, the replacement may be based on a prediction that the second model may perform better than the primary model under the current market conditions. In embodiments of the invention, the prediction may be based at least in parts on machine learning, historical advertising performance data 130, historical event data, and real-time event data 160.”).
As per Claim 4, Simmons teaches a method as provided in claim 1 above. Simmons teaches wherein the hierarchy data structure separates the dataset corresponding to the entity into at least two categories based on a type of information in the dataset (para. 0062, “Contextual data… may be stored as a categorization metadata relating to publisher's content. In an example, such categorization metadata may record that a first publisher's website is related to financial content, and a second publisher's content is predominantly sports-related.”; para. 0089, “contextual category of a medium for advertising. For example, a contextualizer may analyze web content to determine whether a web page contains content about sports, finance, or some other topic. This information may be used as an input to the learning system 138, to refine which types of pages on which ads will appear.”; also see para. 0150 and 0201).
As per Claim 5, Simmons teaches a method as provided in claim 1 above. Simmons teaches combining, by the one or more processors, an output of the first featurization process with an output of the second featurization process to generate the feature set (para. 0170, “in an embodiment of the invention, the real-time bidding machine facility 142 may use a primary model for predicting an economic valuation of each of a plurality of available web publishable advertisement placements based in part on past performance and prices of similar advertisement placements. The real-time bidding machine facility 142 may also use a second model for predicting an economic valuation of each of the plurality of web publishable advertisement placements. After predicting the economic valuations using both the primary model and the second model, the real-time bidding machine facility 142 may compare the valuations produced by the primary model and the second model to determine a preference between the primary model and the second model. In an embodiment of the invention, the comparison of the valuations may include retrospectively comparing the extent to which the models reflect actual economic performance of advertisements. Further, in an embodiment of the invention, the primary model may be an active model responding to purchase requests. The purchase request may be a time limited purchase request. In an embodiment of the invention, the second model may replace the primary model as the active model responding to purchase requests. Further, the replacement may be based on a prediction that the second model may perform better than the primary model under the current market conditions. In embodiments of the invention, the prediction may be based at least in parts on machine learning, historical advertising performance data 130, historical event data, and real-time event data 160.”).
As per Claim 6, Simmons teaches a method as provided in claim 1 above. Simmons teaches wherein the first subset of the dataset corresponds the entity and the second subset of the dataset corresponds to the entity and a sub-entity of the entity (para. 0006, “In embodiments, the third party dataset may include data relating to users of advertising content, contextual data relating to the plurality of available placements, and financial data relating to historical advertisement impressions. The contextual data may derive from a contextualizer service that may be associated with the machine learning facility. In embodiments, data relating to users of advertising content may include demographic data, transaction data, and advertisement conversion data. In embodiments, the economic valuation model may be based in part on real time event data, historic event data, user data, third-party commercial data, advertiser data, and advertising agency data.”).
As per Claim 7, Simmons teaches a method as provided in claim 1 above. Simmons teaches wherein the first featurization process or the second featurization process comprises an aggregation or replication process (para. 00164, 0174, “Further, in embodiments of the invention, the learning machine facility 138 may utilize various types of algorithms to refine the economic valuation models of the real-time bidding machine facility 142. The algorithms may include, without any limitations, decision tree learning, association rule learning, artificial neural networks, genetic programming, inductive logic programming, support vector machines, clustering, Bayesian networks, and reinforcement learning. ”).
As per Claim 8, Simmons teaches a method as provided in claim 1 above. Simmons teaches wherein the machine-learning model comprises a plurality of models, and wherein executing the machine-learning model comprises providing, by the one or more processors, the feature set as input to the plurality of models (para. 0164, 0174, “Further, in embodiments of the invention, the learning machine facility 138 may utilize various types of algorithms to refine the economic valuation models of the real-time bidding machine facility 142. The algorithms may include, without any limitations, decision tree learning, association rule learning, artificial neural networks, genetic programming, inductive logic programming, support vector machines, clustering, Bayesian networks, and reinforcement learning. ”; para. 0150, “Further, in embodiments of the invention, a data integration facility 134 and the contextualizer service facility 132 may be associated with the analytics platform facility 114 and the one or more databases. The data integration facility 134 may facilitate the integration of different types of data from one or more databases into the analytics platform facility 114. The contextualizer service facility 132 may identify the contextual category of a medium for advertising and/or publisher content, website, or other publisher ad context. In an example, a contextualizer may analyze web content to determine whether a web page contains content about sports, finance, or some other topic. This information may be used as an input to the learning machine facility in order to identify the relevant publishers and/or web pages where ads may appear. In another embodiment, the location of the ad on the publisher 112 web page may be determined based on the information. In an embodiment of the invention, the contextualizer service facility 132 may also be associated with the real-time bidding machine facility 142 and/or with the one or more databases.”; para. 0005-0006, “The machine learning may be based at least in part on a third party dataset… the third party dataset may include data relating to users of advertising content, contextual data relating to the plurality of available placements, and financial data relating to historical advertisement impressions. The contextual data may derive from a contextualizer service that may be associated with the machine learning facility. In embodiments, data relating to users of advertising content may include demographic data, transaction data, and advertisement conversion data. In embodiments, the economic valuation model may be based in part on real time event data, historic event data, user data, third-party commercial data, advertiser data, and advertising agency data”; also see para. 0055-0056, regarding the databases providing add related data and historical event data to the machine learning).
As per Claim 9, Simmons teaches a method as provided in claim 1 above. Simmons teaches wherein the plurality of models comprises at least one clustering model (para. 00164, 0174, “Further, in embodiments of the invention, the learning machine facility 138 may utilize various types of algorithms to refine the economic valuation models of the real-time bidding machine facility 142. The algorithms may include, without any limitations, decision tree learning, association rule learning, artificial neural networks, genetic programming, inductive logic programming, support vector machines, clustering, Bayesian networks, and reinforcement learning. ”).
As per Claim 10, Simmons teaches a method as provided in claim 1 above. Simmons teaches updating, by the one or more processors, the machine-learning model responsive to generating the feature set (para. 0127, “This data may be prepared for machine learning training by labeling each ad impression with an indicator of whether or not it leads to a click or action. Alternative machine learning algorithms may be trained on the labeled data. ”; also see para. 0137).
Response to Arguments
Applicant’s arguments are fully considered and would not overcome all of the rejections. Details are provided below.
Arguments on rejections under 35 U.S.C. 101:
Applicants argued that the claims do not recite an abstract idea. Examiner respectfully disagrees.
Applicant’s claimed invention is directed to abstract idea of mental processes (observing and evaluating data regarding entities to make a judgement/opinion on generating feature sets), and certain methods of organizing human activities, specifically fundamental economic practice (generating feature sets from entity/business data possibly for managing future acts of the entity better), managing personal behavior and interactions (following rules and instructions to generate and utilize feature sets from data associated to entities).
Applicants argued that the claims are eligible based on improvement to computer systems. Examiner respectfully disagrees.
Additional elements are recited at a high level of generality and are not improved by this invention. They are merely “applying” and/or “generally linking” the abstract idea to a particular technological environment.
Arguments on rejections under 35 U.S.C. 102:
Applicant’s arguments are geared towards newly added/amended limitations that are considered for the first time above in the rejection section of this Office action to which the applicants may refer for further clarification.
Conclusion
Additional prior not relied upon includes:
Murray et al (US 20190102438 A1), regarding “The metadata produced by profile engine 1426 can be provided to the recommendation engine 1408 to generate one or more transform recommendations. The entities that match an identified pattern of the data can be used to enrich the data with those entities identified by classification determined using knowledge service 1410. In some embodiments, the data to the identified patterns (e.g., city and state) may be provided to knowledge service 1410 to obtain, from a knowledge source 1440, entities that match the identified patterns. For example, knowledge service 1410 may be invoked calling a routine (e.g., getCities( )and getStates( )) corresponding to the identified patterns to receive entity information. The information received from knowledge service 1410 may include a list (e.g., canonical list) of entities that have properly spelled information (e.g., properly spelled cities and states) for the entities. Entity information corresponding to matching entities obtained from knowledge service 1410 can be used to enrich data, e.g., normalize the data, repair the data, and/or augment the data.”;
Hu et al (CN 110858377 A), regarding “Specifically, the user behavior data may comprise at least one of data characteristic information and the behaviour attribute corresponding to each data. wherein the characteristic information of the data may be further expanded feature vector corresponding to data. can comprise all values corresponding to the characteristic feature vector. value to facilitate subsequent calculation, corresponding to each feature item can be binary elements, i.e., the value of the element is 0 or 1, the feature vector so that data corresponding to into a binary vector. For example, the feature vector corresponding to a commodity is (A1, A2), assuming A1 is the colour characteristic item corresponding to the binary elements, the value of black is 0, the value of white is 1, A2 is binary element corresponding to the brand characteristic item, brand C takes a value of 1, the value of the non-brand C 0; if the colour of the product is white, then the feature vector in the value of A1 is 1, the brand of the product is brand B, the characteristic vector A2 takes a value of 0.
In one scheme capable of realizing, in the S2 "to the user behavior data for data analysis to obtain the user individual information" specifically comprises: the behaviour data as the input parameter of the local calculation model, executing the calculation model to obtain the user individual information.
In the above embodiment, user personal information (also called user preferences) is based on user behavior data generated by the said page operation determined to not reference history operation behaviour data of the user. Generally speaking, based on historical behavior data and real-time behaviour data, can more accurately analyze the personality information of the user. Therefore, in order to improve analysis accuracy of the personality information, in specific implementation, the historical behaviour data and real-time behavior data are used as input parameter calculation model, to execute the calculation model to obtain the user individual information.”.
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MEHMET YESILDAG whose telephone number is (571)272-3257. The examiner can normally be reached M-F 8:30 am - 5:00 pm.
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, Jerry O'Connor can be reached at (571) 272-6787. 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.
/MEHMET YESILDAG/Primary Examiner, Art Unit 3624