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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 02/19/2026 has been entered.
Response to Arguments
Applicant’s arguments filed 12/22/2025 on pages 11-20 of Remarks regarding the rejection under 35 U.S.C. 101 with respect to claims 1, 4-5 and 7-20 have been fully considered but they are not persuasive.
Beginning on page 12 Applicant asserts that under 101 Step 2A Prong One the claims are not directed to an abstract idea because the amended independent claim 1 provides an improvement to event forecasting, prediction and anomaly detection. However, Examiner respectfully disagrees. MPEP 2106.04(a)(2)(III)(c) talks about mental processes on a generic computer. Also, see MPEP 2106.04(d) and 2106.05(f). The above mentioned sections of the MPEP set forth that a claim may recite a mental process even with the use of a generic computer. In this instance, generic computing components mentioned such as a processor and machine learning model are used along side with bipartite graphs for event forecasting, prediction and anomaly detection.
Beginning on page 18 Applicant asserts that under 101 Step 2A Prong Two and 101 Step 2B that the features of amended independent claim 1 apply the judicial exception by, at least, improving the technology/technical field of computer-driven event forecasting and anomaly detection. However, Examiner respectfully disagrees. MPEP 2106.04(d)(1) talks about a claim reciting a judicial exception is not directed to the judicial exception if it also recites additional elements demonstrating that the claim as a whole integrates the exception into a practical application. MPEP 2106.05(a) talks about the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements. Improving an abstract element is not sufficient to integrate into a practical application. See updated rejection below.
Applicant’s arguments on pages 20-24 regarding the rejection under 35 U.S.C. 103 with respect to claims 1, 4-5 and 7-20 have been fully considered but are moot in view of new references Hu, Durairaj and Li.
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, 4-5 and 7-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding Claim 1,
Claim 1 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 1 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“detect a plurality of motifs representing a plurality of events in the dataset of data instances using a matrix profile-based motif detection technique, each motif of the plurality of motifs comprising a signal pattern representing an event, and the plurality of motifs being associated with a set of most frequently repeated patterns in the dataset of data instances;”
“generate a bipartite graph representation of the plurality of motifs in a time sequence, wherein the bipartite graph representation comprises a first set of nodes associated with a plurality of intervals of time series data and a second set of nodes associated with the plurality of events represented by the plurality of motifs, and wherein, when generating the bipartite graph representation of the plurality of motifs, [the at least one processor is programmed or configured to:]”
“determine a plurality of features associated with the second set of nodes based on each event of the plurality of events represented by the plurality of motifs, and”
“determine a plurality of features associated with the second set of nodes representing a plurality of edges of the bipartite graph representation based on a time at which each event of the plurality of events represented by the plurality of motifs occurred in the time sequence;”
“generate a machine-learning model based on the bipartite graph representation of the plurality of motifs in the time sequence, wherein the machine-learning model is configured to provide an output comprising a motif predicted to occur during a specified time interval;”
“generate an anomaly score associated with the specified time interval based on time-series data from the specified time interval and a projection of the motif predicted to occur; and”
“determine a motif predicted to occur in a time interval [using the machine-learning model;]”
“detect an anomaly in the specified time interval based at least partly on the anomaly score; and”
“determine whether the motif predicted to during the specified time interval corresponds to ground truth data indicating whether the motif did occur during the specified time interval; and”
“update weight parameters of the machine-learning model based on determining whether the motif predicted to occur during the specified time interval corresponds to ground truth data indicating whether the event did occur at the specified time interval.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., detecting, generating, determining, updating). The above limitations in the context of this claim encompass, inter alia, detecting motifs, generating graphs, determining features, generating an anomaly score, detecting an anomaly, updating weight parameters (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: The judicial exceptions are not integrated into a practical application.
The limitations:
“at least one processor programmed or configured to:”
“[determine a motif predicted to occur in a time interval] using the machine-learning model;”
“train the machine-learning model, wherein, when training the machine-learning model, the at least one processor is programmed or configured to:”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor and machine learning model (e.g., by using these elements as tools).
The limitations:
“receive a dataset of data instances, wherein each data instance comprises a time series of data points;”
As drafted, amount to insignificant extra-solution activities, which do not integrate a judicial exception into a practical application. For example, the additional elements of
"receiving a dataset" amount to mere data gathering and data storage, respectively, which
are insignificant extra-solution activities that do not integrate a judicial exception into a
practical application. See MPEP 2106.05(g).
Step 2B Analysis: The claim does not include additional elements that are sufficient to
amount to significantly more than the judicial exception.
The limitations:
“at least one processor programmed or configured to:”
“[determine a motif predicted to occur in a time interval] using the machine-learning model;”
“train the machine-learning model, wherein, when training the machine-learning model, the at least one processor is programmed or configured to:”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor and machine learning model (e.g., by using these elements as tools).
The limitations:
“receive a dataset of data instances, wherein each data instance comprises a time series of data points;”
As discussed above with respect to integration of the abstract idea into a practical application, all of the additional elements are insignificant extra-solution activities or mere instructions to apply an exception. (i.e., the additional element describes a unit for applying the abstract ideas). Insignificant extra-solution activities and mere instructions to apply an exception cannot provide an inventive concept. Moreover, receiving, communicating, and storing data are insignificant extra-solution activities that are well-understood, routine, and conventional. See MPEP 2106.05(d)(II) ("The courts have recognized the following computer functions as well-understood, routine, and conventional functions ... i. Receiving or transmitting data over a network") (citing OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015)). The claim is not patent eligible.
Regarding Claim 4,
Claim 4 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 4 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“wherein the machine-learning model is configured to provide the output based on an input, and wherein the input comprises one or more time series of data points.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., calculating). The above limitations in the context of this claim encompass, inter alia, calculating an anomaly score (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: Please see the corresponding analysis of Claim 1.
Step 2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 5,
Claim 5 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 5 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“determine a matrix profile score for each data instance of the dataset of data instances; and”
“detect the plurality of motifs representing the plurality of events in the dataset of data instances based on the matrix profile score for each data instance of the dataset of data instances.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., determining). The above limitations in the context of this claim encompass, inter alia, determining and detecting (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: Please see the corresponding analysis of Claim 1.
Step 2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 7,
Claim 7 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 7 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“detect each motif of the plurality of motifs according to a plurality of time intervals in which the plurality of motifs are located using the matrix profile-based motif detection technique; and”
“wherein, when generating the bipartite graph representation of the plurality of motifs in the time sequence, [the at least one processor is programmed or configured to:]”
“generate the bipartite graph representation of the plurality of motifs in the time sequence based on the plurality of time intervals in which the plurality of motifs are located.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., detecting). The above limitations in the context of this claim encompass, inter alia, detecting and generating the bipartite graph (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: The judicial exceptions are not integrated into a practical application.
The limitations:
“the at least one processor is programmed or configured to:”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor (e.g., by using these elements as tools).
Step 2B Analysis:
The limitations:
“the at least one processor is programmed or configured to:”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor (e.g., by using these elements as tools).
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 8,
Claim 8 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 8 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“wherein the bipartite graph representation further comprises at least one residual node associated with a residual error.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., representing). The above limitations in the context of this claim encompass, inter alia, a representation of nodes on a graph (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: Please see the corresponding analysis of Claim 1.
Step 2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 9,
Claim 9 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 9 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“wherein the at least one residual node comprises a first residual node and a second residual node, wherein the first residual node indicates the residual error is larger than a threshold, and wherein the second residual node indicates the residual error is equal to or less than the threshold.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., indicating). The above limitations in the context of this claim encompass, inter alia, nodes indicating residual error (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: Please see the corresponding analysis of Claim 1.
Step 2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 10,
Claim 10 is rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1 Analysis: Claim 10 is directed to a system, i.e., a machine, one of the statutory categories.
Step 2A Prong One Analysis: The limitations:
“[wherein the at least one processor is further programmed or configured to] generate the anomaly score based on at least one of the following:”
“an event forecasting score based on a probability value of at least one signal pattern in the bipartite graph representation,”
“a residual score based on a frequency of change of at least one signal pattern in the bipartite graph representation, or any combination thereof.”
As drafted, under their broadest reasonable interpretation, cover concepts performed in human mind (including an observation, evaluation, judgement, or opinion, e.g., calculating). The above limitations in the context of this claim encompass, inter alia, calculating a score (corresponding to mental processes which can be done mentally or by pen and paper).
Step 2A Prong Two Analysis: The judicial exceptions are not integrated into a practical application.
The limitations:
“wherein the at least one processor is further programmed or configured to”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor (e.g., by using these elements as tools).
Step 2B Analysis:
The limitations:
“wherein the at least one processor is further programmed or configured to”
As drafted, are additional elements that amount to no more than mere instructions to apply the exception for the abstract ideas. See MPEP 2106.05(f). Specifically, they amount to mere instructions to apply the exception using a processor (e.g., by using these elements as tools).
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The claim is not patent eligible.
Regarding Claim 11,
Claim 11 recites a computer-implemented method for performing steps substantially similar to those of claim 1 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 12,
Claim 12 recites a computer-implemented method for performing steps substantially similar to those of claim 5 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 13,
Claim 13 recites a computer-implemented method for performing steps substantially similar to those of claim 7 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 14,
Claim 14 recites a computer-implemented method for performing steps substantially similar to those of claim 9 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 15,
Claim 15 recites a computer-implemented method for performing steps substantially similar to those of claim 10 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 16,
Claim 16 recites a computer program product comprising at least one non-transitory computer-readable medium for performing steps similar of 1 and is rejected with the same rationale, mutatis mutandis, in view of the following additional elements, considered individually and as an ordered combination with the additional elements identified above, failing to integrate the abstract idea into a practical application or amount to significantly more than the abstract idea:
“the computer program product comprising at least one non-transitory computer-readable medium including program instructions that, when executed by at least one processor, cause the at least one processor to:”
This is a recitation of generic computer components to be used in performing the abstract idea, which does not integrate the abstract idea into a practical application or amount to significantly more than the abstract idea. See MPEP 2106.05(f).
Regarding Claim 17,
Claim 17 recites a non-transitory computer-readable medium for performing steps substantially similar to those of claim 5 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 18,
Claim 18 recites a non-transitory computer-readable medium for performing steps substantially similar to those of claim 7 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 19,
Claim 19 recites a non-transitory computer-readable medium for performing steps substantially similar to those of claim 9 and is rejected with the same rationale, mutatis mutandis.
Regarding Claim 20,
Claim 20 recites a non-transitory computer-readable medium for performing steps substantially similar to those of claim 10 and is rejected with the same rationale, mutatis mutandis.
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.
Claims 1, 4, 5, 7, 11, 12, 13, 16, 17 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Hu et al. (Time-Series Event Prediction with Evolutionary State Graph); hereinafter Hu in view of Law et al. (US20200258157A1); hereinafter Law in view of Durairaj et al. (US 20180316704 A1); hereinafter Durairaj and in view of Li et al. (Recommendation as link prediction in bipartite graphs: A graph kernel-based machine learning approach); hereinafter Li
Claim 1 is rejected over Hu, Law, Durairaj and Li.
Regarding claim 1, Hu teaches a system for event forecasting using a graph-based machine-learning model, the system comprising:
at least one processor programmed or configured to: (Hu [Abstract]: “we propose a novel graph neural network model, Evolutionary State Graph Network (EvoNet), to encode the evolutionary state graph for accurate and interpretable time-series event prediction.”)
receive a dataset of data instances, wherein each data instance comprises a time series of data points; (Hu [2 Background and Problem]: “Once the states have been recognized, one can then models each time-series segment X𝑡 as a composition of states–i.e., quantify the recognition weight of each state for a segment to characterize the segment-state associations.”)
determine a motif predicted to occur in a time interval using the machine-learning model; (Hu [3 Evonet Framework]: “we present a novel framework for time-series event prediction. We name the proposed framework Evolutionary State Graph Network (EvoNet), as it transforms the time-series into a dynamic graph based on the states and recognition weights, and constructs a GNN-based neural network to capture significant correlations and improve the ability of event prediction.”)
generate an anomaly score associated with the time interval based on time-series data from the time interval and a projection of the motif predicted to occur; (Hu [page 3-4]: “we conduct statistic analysis related to the constructed evolutionary state graph based on the abnormal anomaly occurs). The distributions of the different graph-level and node-level measurements at different times (before and after anomaly 𝑡) are visualized in Figure 3. From the figure, we can clearly see that when an anomaly occurs, the abnormal graph (red bar) tends to be denser; i.e., the betweenness scores gets lower, while the closeness scores gets higher. Figure 3b presents three typical states and compares their in-degree before and after anomaly 𝑡.”; and [page 6, 4.1 Datasets]: “Our goal is to predict future anomalies (next day) based on records from the past 15 days. In total, we identify around 200K normal cases and 20K abnormal ones.”)
detect an anomaly in the time interval based at least partly on the anomaly score; and (Hu [page 3-4]: “we conduct statistic analysis related to the constructed evolutionary state graph based on the abnormal anomaly occurs). The distributions of the different graph-level and node-level measurements at different times (before and after anomaly 𝑡) are visualized in Figure 3. From the figure, we can clearly see that when an anomaly occurs, the abnormal graph (red bar) tends to be denser; i.e., the betweenness scores gets lower, while the closeness scores gets higher. Figure 3b presents three typical states and compares their in-degree before and after anomaly 𝑡.”)
train the machine-learning model, wherein, when training the machine-learning model, the at least one processor is programmed or configured to:
determine whether the motif predicted to occur during the time interval corresponds to ground truth data indicating whether the motif did occur during the time interval; and (Hu [page 5, Temporal graph propagation]: To learn the parameters 𝜃 of the proposed EvoNet and classifier, we employ an end-to-end framework, based on the Adam optimization algorithm [26] to minimize the cross-entropy loss L as follows: (9) where
Y
^
t+1 ∈ {0,1} is the ground truth that indicating whether a future event will occur.”)
update weight parameters of the machine-learning model based on determining whether the motif predicted to occur during the time interval corresponds to the ground truth data indicating whether the motif did occur at the time interval. (Hu [page 11, A.1 Algorithm Details]: “Finally, the learned representations are fed into an output model for prediction tasks; we use a backpropagation learning algorithm with cross-entropy loss to train the entire networks. More details can be found in Algorithm 1.”; Note: The weight parameters are updated through backpropagation)
Hu does not appear to explicitly teach detect a plurality of motifs representing a plurality of events in the dataset of data instances using a matrix profile-based motif detection technique;
each motif of the plurality of motifs comprising a signal pattern representing an event, and the plurality of motifs being associated with a set of most frequently repeated patterns in the dataset of data instances;
However, Law teaches detect a plurality of motifs representing a plurality of events in the dataset of data instances using a matrix profile-based motif detection technique; (Law [0183]: “The lowest local maxima in the matrix profile values (indicated in FIG. 18 with open triangles) are considered a motif since they represent the pair of nearest neighbor subsequences with the smallest z-normalized Euclidean distance.”; [0115]: “If the user identifies a pattern of interest (such as a pattern (motif) that precedes certain events that may drive investment decisions), the user can provide a name for that pattern at UI element 560 and then save the named pattern to a pattern library 564 using UI element 568.”; and [0067]: “The matrix profile 320 is plotted on the same x-axis as the time-series data 300 and each value of the matrix profile 320 corresponds to a subsequence of length m in the time-series data 300. For illustration, lines 324 and 328 have been drawn at the global minima of the matrix profile 320. The value of the matrix profile 320 at line 324 corresponds to a subsequence of length m beginning in the time-series data 300 at line 324.”)
each motif of the plurality of motifs comprising a signal pattern representing an event, and the plurality of motifs being associated with a set of most frequently repeated patterns in the dataset of data instances; (Law [0168]: “Time series motifs are approximately repeated subsequences found within a longer time series. Being able to say that a subsequence is “approximately repeated” requires that you be able to compare subsequences to each other.”; and [0064]: “Certain patterns may indicate to investors that an event may be more likely than usual. For example, a certain pattern may presage a stock price decrease for a company following a below-expectations earnings report. Other patterns may signal a decrease in confidence in the equity by institutional investors.”)
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the matrix profile-based motif detection technique of Law for computational efficiency (Law [0077]). Hu and Law are analogous art because they both concern analyzing patterns in time-series data.
Hu does not appear to explicitly teach generate a bipartite graph representation of the plurality of motifs in a time sequence, wherein, when generating the bipartite graph representation of the plurality of motifs, the at least one processor is programmed or configured to:
determine a plurality of features representing a plurality of nodes of the bipartite graph representation based on each event of the plurality of events represented by the plurality of motifs, and
determine a plurality of features representing a plurality of edges of the bipartite graph representation based on a time at which each event of the plurality of events represented by the plurality of motifs occurred in the time sequence; and
However, Durairaj teaches generate a bipartite graph representation of the plurality of motifs in a time sequence, wherein, when generating the bipartite graph representation of the plurality of motifs, the at least one processor is programmed or configured to: (Durairaj [0042]: “The event data include time-related data, such as timestamp data, which enables some or all of the event data to be sequenced. The LM security application analyzes the timestamp data to sequence the event data, and creates a data structure which represents an associated time constrained graph. The graph includes nodes and connections between nodes.”)
determine a plurality of features representing a plurality of nodes of the bipartite graph representation based on each event of the plurality of events represented by the plurality of motifs, and (Durairaj [0061]: “Every edge in the bipartite graph connects a user node in the first set to a device node in the second set. In addition, the relationships 230 between the user nodes and the device nodes also represent a time series of events in which the users have interacted (e.g., logged in) with the network devices.”)
determine a plurality of features representing a plurality of edges of the bipartite graph representation based on a time at which each event of the plurality of events represented by the plurality of motifs occurred in the time sequence; and (Durairaj [0061]: “Every edge in the bipartite graph connects a user node in the first set to a device node in the second set. In addition, the relationships 230 between the user nodes and the device nodes also represent a time series of events in which the users have interacted (e.g., logged in) with the network devices.”)
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the bipartite graph representation of Durairaj to effectively detect anomalies based on the relationships between the nodes (Durairaj [0067]). Hu and Durairaj are analogous art because they concern analyzing temporal relationships in graphs to detect anomalies.
Hu does not appear to explicitly teach generate a machine-learning model based on the bipartite graph representation of the plurality of motifs in the time sequence,
However, Li teaches generate a machine-learning model based on the bipartite graph representation of the plurality of motifs in the time sequence, (Li [2.1. Recommendation Algorithmns]: “Collective local features capture the collective characteristics of individual user/item information, such as a user's demographic characteristics, content of interest [32], item's specifications, transaction contexts (environment, time, etc.), and temporal usage patterns (which reflect user's characteristics) … Graph structure can be used to design similarity measures for cross-recommendation. The bipartite user–item graph can also be projected onto a unipartite user/item graph [70] to simplify the graph structure. Furthermore, some researchers construct graphs based on user/item similarities.”; [3.1 A graph kernel-based recommendation framework]: “Fig. 1 shows the four steps in our graph kernel-based framework. 1) In the graph and feature extraction step, we construct the user–item graph from transaction histories. We also extract features describing nodes (users and items) from the data. 2) In the graph kernel construction step, we design a kernel k() on user–item pairs based on their context structure and features. In kernel-based methods, the kernel design is the most essential module for prediction. 3) In the model building step, a classifier is built to separate potential links from impossible links.”)
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the bipartite graph kernel-based approach of Li for improved recommendation performance (Li [5. Conclusions and future directions]). Hu and Li are analogous art because they both concern forecasting using network graphs.
Claim 4 is rejected over Hu, Law, Durairaj and Li with the incorporation of claim 1.
Regarding claim 4, Hu teaches wherein the machine-learning model is configured to provide the output based on an input, and wherein the input comprises one or more time series of data points. (Hu [2 Background and Problem]: “Once the states have been recognized, one can then models each time-series segment X𝑡 as a composition of states–i.e., quantify the recognition weight of each state for a segment to characterize the segment-state associations.”; and [5 Related Work]: “Models in the graph family [4, 15, 16, 22, 27, 42, 46] have been applied to many real-world scenarios, including learning the dynamics of physical systems [5, 40], predicting the chemical properties of molecules [17], predicting traffic on roads [18] and reasoning about knowledge graphs [21], etc..”)
Claim 5 is rejected over Hu, Law, Durairaj and Li with the incorporation of claim 1.
Regarding claim 5, Hu does not appear to explicitly teach when detecting the plurality of motifs representing the plurality of events in the dataset of data instances using the matrix profile-based motif detection technique, the at least one processor is programmed or configured to:
determine a matrix profile score for each data instance of the dataset of data instances; and
detect the plurality of motifs representing the plurality of events in the dataset of data instances based on the matrix profile score for each data instance of the dataset of data instances.
However, Law teaches when detecting the plurality of motifs representing the plurality of events in the dataset of data instances using the matrix profile-based motif detection technique, the at least one processor is programmed or configured to:
determine a matrix profile score for each data instance of the dataset of data instances; and (Law [0168]: “Time series motifs are approximately repeated subsequences found within a longer time series. Being able to say that a subsequence is “approximately repeated” requires that you be able to compare subsequences to each other. In the case of Stumpy, all subsequences within a time series can be compared by computing the pairwise z-normalized Euclidean distances and then storing only the index to its nearest neighbor. This nearest neighbor distance is referred to as the matrix profile and the index to each nearest neighbor within the time series is referred to as the matrix profile index. Luckily, the stump function is configured to take in any time series (with integer or floating point values) and compute the matrix profile along with the matrix profile indices. With the matrix profile and matrix profile indices, finding time series motifs is straightforward.”; Note: The nearest neighbor distance is the matrix profile score.)
detect the plurality of motifs representing the plurality of events in the dataset of data instances based on the matrix profile score for each data instance of the dataset of data instances. (Law [0064]: “Certain patterns may indicate to investors that an event may be more likely than usual. For example, a certain pattern may presage a stock price decrease for a company following a below-expectations earnings report. Other patterns may signal a decrease in confidence in the equity by institutional investors.”)
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the matrix profile-based motif detection technique of Law for computational efficiency (Law [0077]). Hu and Law are analogous art because they both concern analyzing patterns in time-series data.
Claim 7 is rejected over Hu, Law, Durairaj and Li with the incorporation of claim 1.
Regarding claim 7, Hu does not appear to explicitly teach wherein, when detecting the plurality of motifs representing the plurality of events in the dataset of data instances using the matrix profile-based motif detection technique, the at least one processor is programmed or configured to:
detect each motif of the plurality of motifs according to a plurality of time intervals in which the plurality of motifs are located using the matrix profile-based motif detection technique; and
However, Law teaches wherein, when detecting the plurality of motifs representing the plurality of events in the dataset of data instances using the matrix profile-based motif detection technique, the at least one processor is programmed or configured to:
detect each motif of the plurality of motifs according to a plurality of time intervals in which the plurality of motifs are located using the matrix profile-based motif detection technique; and (Law [0067]: “The matrix profile 320 is plotted on the same x-axis as the time-series data 300 and each value of the matrix profile 320 corresponds to a subsequence of length m (time intervals) in the time-series data 300. For illustration, lines 324 and 328 have been drawn at the global minima of the matrix profile 320. The value of the matrix profile 320 at line 324 corresponds to a subsequence of length m beginning in the time-series data 300 at line 324.”)
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the matrix profile-based motif detection technique of Law for computational efficiency (Law [0077]). Hu and Law are analogous art because they both concern analyzing patterns in time-series data.
Claim 11 is rejected over Hu, Law, Durairaj and Li.
The remainder of claim 11 is claim 1 in the form of a method and is rejected for the same reasons as claim 1 stated above.
Dependent claim 12 is claim 5 in the form of a method and is rejected for the same reasons as claim 5 stated above. For the rejection of the limitations specifically pertaining to the method of claim 11, see the rejection of claim 11 above.
Dependent claim 13 is claim 7 in the form of a method and is rejected for the same reasons as claim 7 stated above. For the rejection of the limitations specifically pertaining to the method of claim 11, see the rejection of claim 11 above.
Claim 16 is rejected over Hu, Law, Durairaj and Li.
The remainder of claim 16 is claim 1 in the form of a non-transitory computer-readable medium and is rejected for the same reasons as claim 1 stated above.
Dependent claim 17 is claim 5 in the form of a non-transitory computer-readable medium and is rejected for the same reasons as claim 5 stated above. For the rejection of the limitations specifically pertaining to the non-transitory computer-readable medium of claim 16, see the rejection of claim 16 above.
Dependent claim 18 is claim 7 in the form of a non-transitory computer-readable medium and is rejected for the same reasons as claim 7 stated above. For the rejection of the limitations specifically pertaining to the non-transitory computer-readable medium of claim 16, see the rejection of claim 16 above.
Claims 8, 9, 10, 14, 15, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Hu, Law, Durairaj, Li and Bertiger et al. (US20210194907A1); hereinafter Bertiger
Claim 8 is rejected over Hu, Law, Durairaj, Li and Bertiger with the incorporation of claim 1.
Regarding claim 8, Hu does not appear to explicitly teach wherein the plurality of nodes of the bipartite graph representation further comprise at least one residual node associated with a residual error.
However, Bertiger teaches wherein the plurality of nodes of the bipartite graph representation further comprise at least one residual node associated with a residual error. (Bertiger [0020]: “The predicted links can subsequently be used, e.g., for anomaly detection by comparing the estimated link probabilities against an observed adjacency matrix for the time interval and identifying nodes and corresponding devices exhibiting anomalous behavior (operations 250-280 in FIG. 2).”; [0045] and “Similarly, the absence of a link in the observed network (represented by a 0 in the adjacency matrix for the second time interval) may be considered an anomalous disruption if the corresponding estimated link probability exceeds a certain threshold. Instead of comparing scores at the level of links, it is also possible to score nodes, aggregating for each node across its incoming and outgoing links, and comparing the aggregated node score (residual node) against a specified threshold.”; and [0040]: “Hence, each node has two different latent positions, representing the behavior of the node as source or destination of the link. Note that DASE can also be extended to bipartite graphs.”).
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the temporal link prediction of Bertiger to efficient prediction of a network (Bertiger [0023]). Hu and Bertiger are analogous art because they both concern using graphs to identify anomalies.
Claim 9 is rejected over Hu, Law, Durairaj, Li and Bertiger with the incorporation of claim 1.
Regarding claim 9, Hu does not appear to explicitly teach wherein the at least one residual node comprises a first residual node and a second residual node, wherein the first residual node indicates the residual error is larger than a threshold, and
wherein the second residual node indicates the residual error is equal to or less than the threshold.
However, Bertiger teaches wherein the at least one residual node comprises a first residual node and a second residual node, wherein the first residual node indicates the residual error is larger than a threshold, and (Bertiger [0020]: “Similarly, the absence of a link in the observed network (represented by a 0 in the adjacency matrix for the second time interval) may be considered an anomalous disruption if the corresponding estimated link probability exceeds a certain threshold (first residual node). Instead of comparing scores at the level of links, it is also possible to score nodes, aggregating for each node across its incoming and outgoing links, and comparing the aggregated node score against a specified threshold.”)
wherein the second residual node indicates the residual error is equal to or less than the threshold. (Bertiger [0020]: “the comparison is based on an anomaly criterion, such as a specified probability threshold. For example, links in the observed network (represented by l's in the adjacency matrix for the second time interval) may be deemed anomalous if the corresponding estimated link probabilities fall below a certain threshold (second residual node).”
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the temporal link prediction of Bertiger to efficient prediction of a network (Bertiger [0023]). Hu and Bertiger are analogous art because they both concern using graphs to identify anomalies.
Claim 10 is rejected over Hu, Law, Durairaj, Li and Bertiger with the incorporation of claim 1.
Regarding claim 10, Hu does not appear to explicitly teach wherein the at least one processor is further programmed or configured to calculate an anomaly score based on at least one of the following:
an event forecasting score based on a probability value of at least one signal pattern in the bipartite graph representation,
a residual score based on a frequency of change of at least one signal pattern in the bipartite graph representation, or any combination thereof.
However, Bertiger teaches wherein the at least one processor is further programmed or configured to calculate an anomaly score based on at least one of the following:
an event forecasting score based on a probability value of at least one signal pattern in the bipartite graph representation, (Bertiger [0013]: “Link prediction in accordance with the disclosed embodiments involves generating a time series of adjacency matrices representing the time-dependent network graph during a first time period (which includes a time series of first time intervals), and computing, from that time series of adjacency matrices, estimated link probabilities for connections in the network graph during a second time period (which may include one or more second time intervals). A link probability (herein also “link score”) (event forecasting score) indicates, for any given pair of nodes, the probability that these two nodes are connected (e.g., representing, for a given pair of network devices, the probability that the devices communicate (event) during a specified time interval).” and [0040]: “Hence, each node has two different latent positions, representing the behavior of the node as source or destination of the link. Note that DASE can also be extended to bipartite graphs.”)
a residual score based on a frequency of change of at least one signal pattern in the bipartite graph representation, or any combination thereof. (Bertiger [0021]: “predictions are made for multiple second time intervals within a second time period by extrapolating by varying lengths of time beyond the first time period (e.g., using different weighting factors in the computations for different second time intervals). In general, the network behavior predicted based on observations during a time series of first time intervals within a first time period is compared against observed network behavior during a second time interval outside the first time period”; Note: The comparison between predicted network behavior and observed network behavior is the residual score based on the frequency of change over time.).
It would have been obvious before the effective filing date to combine the time-series event prediction graph of Hu with the temporal link prediction of Bertiger to efficient prediction of a network (Bertiger [0023]). Hu and Bertiger are analogous art because they both concern using graphs to identify anomalies.
Dependent claim 14 is claim 9 in the form of a method and is rejected for the same reasons as claim 9 stated above. For the rejection of the limitations specifically pertaining to the method of claim 11, see the rejection of claim 11 above.
Dependent claim 15 is claim 10 in the form of a method and is rejected for the same reasons as claim 10 stated above. For the rejection of the limitations specifically pertaining to the method of claim 11, see the rejection of claim 11 above.
Dependent claim 19 is claim 9 in the form of a non-transitory computer-readable medium and is rejected for the same reasons as claim 9 stated above. For the rejection of the limitations specifically pertaining to the non-transitory computer-readable medium of claim 16, see the rejection of claim 16 above.
Dependent claim 20 is claim 10 in the form of a non-transitory computer-readable medium and is rejected for the same reasons as claim 10 stated above. For the rejection of the limitations specifically pertaining to the non-transitory computer-readable medium of claim 16, see the rejection of claim 16 above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
NPL: Besta, Maciej et al. “Motif Prediction with Graph Neural Networks.” 2021
Any inquiry concerning this communication or earlier communications from the examiner should be directed to DAVID H TRAN whose telephone number is (703)756-1525. The examiner can normally be reached M-F 9:30 am - 5:30 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, Viker Lamardo can be reached on (571) 270-5871. 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.
/DAVID H TRAN/Examiner, Art Unit 2147
/VIKER A LAMARDO/Supervisory Patent Examiner, Art Unit 2147
Prosecution Insights