DETAILED ACTION
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 .
Priority
Applicant claims the benefit of prior-filed National Stage Application filed under 35 U.S.C. §371, International Application No. PCT/CN2022/125693, filed on October 17, 2022, and Chinese Application No. 202111243147.X, filed on October 25, 2021, which is acknowledged.
Drawings
The drawings were received on 12/20/2023. These drawings are acceptable.
Specification
The substitute specification filed 12/20/2023 has been entered.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on the following date(s): 12/20/2023 has been considered by the examiner.
Claim Rejections - 35 USC § 102
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.
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 13 and 14 are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Krishnan et al. (US 12566974, hereinafter ‘Krish’).
Regarding independent claim 1, Krish teaches a knowledge graph processing method, comprising: selecting several nodes and their edges from a shared knowledge graph based on one or more entity types involved in a target service domain, to obtain a target subgraph, wherein the shared knowledge graph is obtained by fusing knowledge graphs of one or more service domains; (68:31-55: In these ways, non-limiting embodiments or aspects of the present disclosure may enable layered inference to support multi-task reasoning, counterfactual updates to a knowledge graph, and/or seamless incorporation of task-specific models [a knowledge graph processing method, comprising: selecting several nodes and their edges from a shared knowledge graph based on one or more entity types involved in a target service domain, to obtain a target subgraph,]. For example, grounding task-specific inferences on a shared base knowledge graph, with intermediate layers enabling incorporation of projections and aggregations of base links, may enable stochastic link inferences for reasoning across multiple related tasks or models [wherein the shared knowledge graph is obtained by fusing knowledge graphs of one or more service domains;]. For example, understanding what data points are not well explained by the existing knowledge space and pushing data driven updates to guide future tasks using an inherently modular approach with intermediate layers forming repositories of task-model based insights may enable the notion of counterfactual examples, in task-model predictions, as updates to a knowledge graph. As an example, a joint approach to incorporate base knowledge graph links and task models to obtain new stochastic links may start with base edges and graph nodes and embed the base edges and graph nodes into a pre-trained latent knowledge space [selecting several nodes and their edges from a shared knowledge graph based on one or more entity types involved in a target service domain, to obtain a target subgraph], evaluate the predictions of a task-specific model against this knowledge space to detect novel counterfactuals, and make updates accordingly in the knowledge space, with the transition to a continuous space enabling utilization of the knowledge graph seamlessly with any differentiable task model. Examiner notes the target subgraph as the subset of associated entities associated with an inference task.)
processing the target subgraph to extract one or more graph features, wherein the graph feature comprises some or all of the following: a node representation vector, an edge representation vector, a graph structure feature, a semantic feature of graph text information, and a graph rule feature; (in 68: 59-67: Performance of machine learning and AI models may be heavily dependent on a choice of features (e.g., data representations, etc.) on which the models are applied. Embeddings may include low dimensional, learned continuous vector representations of data [processing the target subgraph to extract one or more graph features, wherein the graph feature comprises some or all of the following: a node representation vector, an edge representation vector] that make extraction of useful information easier when building classifiers and/or other prediction models. Embedding/representation learning may be particularly useful when there is little to no metadata and label information. And in 69:39-61 : As shown in FIG. 5, at step 502, process 500 includes obtaining a graph including a first layer. For example, transaction service provider system 108 may obtain a graph including a first layer (e.g., a base knowledge graph, a fact layer of a multi-layer knowledge graph, etc.) including a plurality of first edges and a plurality of first nodes for the plurality of first edges. As an example, the plurality of first nodes may be associated with a plurality of first entities (e.g., the plurality of first entities, entity attributes of the plurality of entities, etc.), and the plurality of first edges may be associated with a plurality of first relationships between the plurality of first entities (e.g., between entities, between entities and entity attributes of the entities, etc.). In such an example, the plurality of first nodes may include sets of independent entities and/or entity attributes [a graph structure feature,], and the plurality of first edges may include a set of sets of relations where each set of relations is the collection of all links of a specific link type that connects two fixed entity types [a graph structure feature, a semantic feature of graph text information]. For example, the plurality of first nodes may include first entities of a plurality of different entity types and/or entity attributes of a plurality of different entity attribute types, and/or the plurality of edges may include first edges of a plurality of different relation types [a graph structure feature, a semantic feature of graph text information].)
and providing the graph feature to a target data processing task of the target service domain, wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature, so as to implement the target data processing task. (in 97:36-51: Example experimental analyses on diverse multi-domain datasets are now discussed, in which the experimental results on two public datasets show that non-limiting embodiments or aspects of the present disclosure may effectively integrate task models with knowledge graph embedding [providing the graph feature to a target data processing task of the target service domain, … so as to implement the target data processing task] and permit for model-to-graph knowledge transfer, graph-to-model knowledge transfer, and model-to-model via graph knowledge transfer […wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature, so as to implement the target data processing task]. First, counterfactual enrichment [wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature] with effective task models is shown to significantly improve the quality of node embeddings for modalities with sparse connections by evaluating the updated embeddings on the held-out link completion task. Next, co-training a context-aware neural recommendation model with the knowledge graph [providing the graph feature to a target data processing task of the target service domain, wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature, so as to implement the target data processing task] is shown to lead to simultaneous embedding updates and better model performance for nodes with lower degrees )
Regarding independent claims 13 and 14, Krish teaches a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to: … (in 63:4-27: Device 200 may perform one or more processes described herein. Device 200 may perform these processes based on processor 204 (e.g., a central processing unit (CPU), a graphics processing unit (GPU), etc.) executing software instructions stored by a computer-readable medium, such as memory 206 and/or storage component 208 [a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to]. A computer-readable medium (e.g., a non-transitory computer-readable medium) is defined herein as a non-transitory memory device. A memory device includes memory space located inside of a single physical storage device or memory space spread across multiple physical storage devices. Software instructions may be read into memory 206 and/or storage component 208 from another computer-readable medium or from another device via communication interface 214. When executed, software instructions stored in memory 206 and/or storage component 208 may cause processor 204 to perform one or more processes described herein. Additionally, or alternatively, hardwired circuitry may be used in place of or in combination with software instructions to perform one or more processes described herein. Thus, embodiments or aspects described herein are not limited to any specific combination of hardware circuitry and software [a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to].)
The remaining limitations of claims 13 and 14 are similar to claim 1 limitations and are rejected under the same rationale.
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, 6, 13 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Sztyler et al. (US 20240095805, hereinafter ‘Tyler’) in view of Kim (US 20220284309).
Regarding independent claim 1, Tyler teaches a knowledge graph processing method, comprising: selecting several nodes and their edges from a shared knowledge graph based on one or more entity types involved in a target service domain, to obtain a target subgraph, wherein the shared knowledge graph is obtained by fusing knowledge graphs of one or more service domains; (in [0040] Everything which is logged/observed (e.g. through a sensor network) by the AI unit 104 is transformed into a Knowledge Graph (KG), i.e., a set of triples. In this regard, the AI unit 104 may be seen as an information extraction pipeline that (continuously or periodically) extracts new triples. This set of triples represents entities and objects, relation between them, and corresponding attributes. According to some embodiments, the AI unit 104 may be configured to store the information found in a Temporal Knowledge Graph (TKG), i.e. a KG that also contains temporal facts indicating relationships among entities and objects at different times. Embodiments of the invention combine (i.e. take into account) the time-dependent triples with a list of actions [selecting several nodes and their edges from a shared knowledge graph based on one or more entity types involved in a target service domain, to obtain a target subgraph, wherein the shared knowledge graph is obtained by fusing knowledge graphs of one or more service domains] and a scope of desired outcomes [to obtain a target subgraph] (i.e. States of Interest) to compute recommendations, as will be described in more detail below…)
processing the target subgraph to extract one or more graph features, wherein the graph feature comprises some or all of the following: a node representation vector, an edge representation vector, a graph structure feature, a semantic feature of graph text information, and a graph rule feature; (in [0040] Everything which is logged/observed (e.g. through a sensor network) by the AI unit 104 is transformed into a Knowledge Graph (KG), i.e., a set of triples. In this regard, the AI unit 104 may be seen as an information extraction pipeline that (continuously or periodically) extracts new triples. This set of triples represents entities and objects, relation between them [a semantic feature of graph text information], and corresponding attributes. According to some embodiments, the AI unit 104 may be configured to store the information found in a Temporal Knowledge Graph (TKG), i.e. a KG that also contains temporal facts indicating relationships among entities and objects at different times [processing the target subgraph to extract one or more graph features]. Embodiments of the invention combine (i.e. take into account) the time-dependent triples with a list of actions and a scope of desired outcomes (i.e. States of Interest) to compute recommendations, as will be described in more detail below. It should be noted that the success of the recommendations depends on how well it is possible to observe/read the domain of interest 106 and convert it into a Knowledge Graph [processing the target subgraph to extract one or more graph features], and how well the graphs are classified by the Graph Analyzer 110 (see below). Hence, the richness of the knowledge base and the neural network architecture of the individual components are important aspects and, as a consequence, the recommender system will have to be carefully tuned for each particular use case… [0041] A prediction component, namely the Future-Link-Entity prediction module 108 (hereinafter briefly denoted FLE prediction module), is configured to take from the AI unit 104 i) the newest KG (hereinafter denoted KG.sub.t, where t is a timestamp that refers to the present), ii) prediction results from the past, and iii) known knowledge from the database as input. The prediction results from the past (as mentioned at ii)) may be provided as a set of pairs of KGs [processing the target subgraph to extract one or more graph features, wherein the graph feature comprises some or all of the following: ] (G={(KG.sub.t.sup.t1, KG.sub.t+x.sup.t1) . . . (KG.sub.t.sup.tn, KG.sub.t+x.sup.tn)}), where x describes a step in time, t+x is a timestamp that refers to the future (t+x>t), t1 . . . tn are timestamps that indicated when the respective pair was processed…)
and providing the graph feature to a target data processing task of the target service domain, wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature, so as to implement the target data processing task. (in [0044] The graphs that were analyzed in the past (i.e. G={(KG.sub.t.sup.t1, KG.sub.t+x.sup.t1) . . . (KG.sub.t.sup.tn, KG.sub.t+x.sup.tn)}, as introduced above, with the corresponding weight matrices, and classification results) can be used as training data [providing the graph feature to a target data processing task of the target service domain, wherein the graph feature is used to serve as an input feature of the target data processing task together with a task customization feature, so as to implement the target data processing task]. The GA module 110 [providing the graph feature to a target data processing task of the target service domain] may be implemented as a neural network whose weights [with a task customization feature] are trained with stochastic gradient descent (SGD) using the known knowledge. The GA module 110 classifies the initial input and the output of the FLE prediction module 108, i.e., KG.sub.t and KG.sub.t+x…. [0050] All of these KGs are analyzed by the GA module 110, i.e., the corresponding weight matrix w and the classification results m are computed, as described already above. Across all actions, the comparison of the corresponding matrices allows to extract the strength of the influence of the action. Hence, by comparing the corresponding matrices it is possible to identify the action [with a task customization feature] which influences the future the most in respect of the interest of the user 102. As a result, the system provides a time-dependent recommendation [so as to implement the target data processing task], i.e., when to act (i.e., now, in the future, never), what should be done (i.e. which action), and how the action influences the development of the KG (i.e., how the future is changed).)
Additionally, Kim teaches processing the target subgraph to extract one or more graph features, wherein the graph feature comprises some or all of the following: a node representation vector, an edge representation vector, a graph structure feature, a semantic feature of graph text information, and a graph rule feature, in [0002] A knowledge graph (or KG) is a collection of machine-readable descriptions of interlinked entities including, for example, real-world objects, events, situations, etc. Machine-learning and artificial intelligence systems may be designed or trained to use KGs for backgrounds knowledge and, thereby, enabling a more accurate interpretation of text and speech data. Due to practical limitations, KGs are often constructed to support only domain-specific applications and are designed & constructed according to a framework defined by experts in that domain. However, in some implementations, it may be advantageous to utilize a KG that more comprehensively covers an entire domain of interest or multiple domains. Accordingly, in some implementations, the systems and methods described herein provide mechanisms for aligning and enriching multiple different KGs by identifying and linking correspondences among entities in each separate KG. In some implementations, the systems and methods described herein utilize a mechanism of subgraph typing to facilitate alignment and enrichment of KGs. A subgraph type is a synthetic category or class constructed using a summary of nodes and the nearby edges to capture a characterization of the structures and semantics of the entities in a KG [a semantic feature of graph text information]. The systems and methods described herein create subgraph types for all the nodes in a KG representing entities and then effectively utilizes the subgraph types to handle corner cases that arise during the enrichment process, such as disambiguating mappings between entities labeled with polysemous terms.
Kim and Tyler are analogous art because both involve developing information retrieval and processing techniques using machine learning systems and algorithms.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of the prior art of developing information retrieval and processing techniques for combining data from multiple different knowledge graphs for use in machine-learning systems, as disclosed by Kim with the method of developing information retrieval and processing techniques for machine learning task, as disclosed by Tyler.
One of ordinary skill in the arts would have been motivated to combine the methods disclosed by Kim and Tyler as noted above. Doing so allowing for developing knowledge graphs (KGs) utilize that more comprehensively cover an entire domain of interest or multiple domains; and for providing mechanisms for aligning and enriching multiple different KGs by identifying and linking correspondences among entities in each separate KG, (Kim, 0002).
Regarding claim 2, the rejection of claim 1 is incorporated and Tyler in combination with Kim teaches the method of method according to claim 1, wherein the graph structure feature comprises one or more of the following: degree information, a PageRank value, a node clustering coefficient, closeness centrality, eigenvector centrality, a common neighbor indicator, a Katz indicator, and random walk similarity. (in [0023] According to an embodiment, the graph analyzer may be configured to characterize present and expected situations by analyzing the temporal Knowledge Graphs for the different points in time. For performing this tasks, the graph analyzer may include a neural network that—given i) a set of already classified pairs (present and future) of TKGs as training data, ii) States of Interest defined by a human (e.g., expressed in form of a set of classification labels), and iii) the TKG which needs to be classified—calculates as outcome a confidence score for each State of Interest [wherein the graph structure feature comprises one or more of the following: degree information] and a weight matrix that reflects the difference between the present TKG and the respective future TKG [closeness centrality]…)
Regarding claim 3, the rejection of claim 1 is incorporated and Tyler in combination with Kim teaches the method of method according to claim 1, further comprising: recalling several candidate nodes from the shared knowledge graph based on the target data processing task, wherein the candidate node is a processing object of the target data processing task; (in , [0035] a Diachronic Analyzer module 112, which simulates possible future scenarios and analyses how certain actions would influence the KGs (present and future) [further comprising: recalling several candidate nodes from the shared knowledge graph based on the target data processing task, wherein the candidate node is a processing object of the target data processing task], [0036] an evaluation module 114 to evaluate the created future scenario KGs and to provide a ranked list of actions, effects and outcomes, and [0037] an explanation module 116 to provide those actions as well as explanations to the user 102, and to interact with the user 102… [0053] Further, according to an embodiment, the evaluation module 114 may also compare the KGs across the pairs and compute the respective weight matrix. Based on the differences between the weight matrices and the confidence scores, the evaluation module 114 computes which action is most effective to turn the KG into the State of Interest…)
wherein a recall manner comprises: querying the shared knowledge graph based on a retrieval condition to obtain the candidate node, or obtaining the candidate node from the shared knowledge graph through vector retrieval based on a target vector. (in [0040] Everything which is logged/observed (e.g. through a sensor network) by the AI unit 104 is transformed into a Knowledge Graph (KG), i.e., a set of triples. In this regard, the AI unit 104 may be seen as an information extraction pipeline that (continuously or periodically) extracts new triples. This set of triples represents entities and objects, relation between them, and corresponding attributes. According to some embodiments, the AI unit 104 may be configured to store the information found in a Temporal Knowledge Graph (TKG), i.e. a KG that also contains temporal facts [wherein a recall manner comprises: querying the shared knowledge graph based on a retrieval condition to obtain the candidate node time as the condition for querying the shared knowledge graph] indicating relationships among entities and objects at different times [wherein a recall manner comprises: querying the shared knowledge graph based on a retrieval condition to obtain the candidate node]. Embodiments of the invention combine (i.e. take into account) the time-dependent triples with a list of actions and a scope of desired outcomes (i.e. States of Interest) to compute recommendations, as will be described in more detail below. It should be noted that the success of the recommendations depends on how well it is possible to observe/read the domain of interest 106 and convert it into a Knowledge Graph, and how well the graphs are classified by the Graph Analyzer 110 (see below)…)
Regarding claim 4, the rejection of claim 1 is incorporated and Tyler in combination with Kim teaches the method of method according to claim 1, wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain, (in [0040] Everything which is logged/observed (e.g. through a sensor network) by the AI unit 104 is transformed into a Knowledge Graph (KG), i.e., a set of triples. In this regard, the AI unit 104 may be seen as an information extraction pipeline that (continuously or periodically) extracts new triples. This set of triples represents entities and objects [wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain], relation between them, and corresponding attributes. According to some embodiments, the AI unit 104 may be configured to store the information found in a Temporal Knowledge Graph (TKG), i.e. a KG that also contains temporal facts indicating relationships among entities and objects at different times. Embodiments of the invention combine (i.e. take into account) the time-dependent triples with a list of actions and a scope of desired outcomes (i.e. States of Interest) [wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain] to compute recommendations, as will be described in more detail below. It should be noted that the success of the recommendations depends on how well it is possible to observe/read the domain of interest 106 and convert it into a Knowledge Graph, and how well the graphs are classified by the Graph Analyzer 110 (see below)...)
to obtain the target subgraph further comprises: obtaining a macro feature of the target subgraph, wherein the macro feature comprises one or more of the following: a quantity of entities, degree distribution of a graph, connectivity distribution of a graph, and a data quality score of a graph; (in [0045] According to an embodiment, the GA module 110 may further compute the difference between the respective two graphs via a weight matrix (w.sub.|KG_(t)−KG_(t+x)|), which reflects the actual changes between the two graphs. As a result, the GA module 110 returns the weight matrix and for each value in the list of interest c∈C (i.e., for each possible class) a confidence score P [wherein the macro feature comprises one or more of the following: a quantity of entities, ] for the respective KG [to obtain the target subgraph further comprises: obtaining a macro feature of the target subgraph] (i.e. classification result m={(c, k, P(c|k))|∀c∈C, ∀k∈{KG.sub.t, KG.sub.t+x}})… And in [0060] 3. Characterization, by Graph Analyzer 110, of given and expected situations by analyzing the TKG for all time steps. The Graph Analyzer 110 may include a neural network that is given a set of classified pairs (present and future) of TKGs as training data, the States of Interest defined by a human (class labels), and the TKG which needs to be classified. The outcome is a confidence score for each State of Interest [wherein the macro feature comprises one or more of the following: a quantity of entities, ] and a weight matrix [a data quality score of a graph] that reflects the difference between the present TKG and the respective future TKG.)
and determining, based on the macro feature, whether the target subgraph satisfies a requirement, and upon determining that the target subgraph does not satisfy the requirement, modifying the target subgraph or re-obtaining a target subgraph from the shared knowledge graph. (in [0053] Further, according to an embodiment, the evaluation module 114 may also compare the KGs across the pairs and compute the respective weight matrix. Based on the differences between the weight matrices and the confidence scores, the evaluation module 114 computes which action is most effective to turn the KG into the State of Interest… And in [0025] According to an embodiment, the recommender system comprises an evaluation module that is configured to compute recommendations of actions (together with their recommended time of execution) based on the graph classification results received from the diachronic analyser. In addition, it may also take into account a model uncertainty of the graph classification results, a difference to alternative scenarios and/or a respective cost associated with each of the actions. Specifically, the evaluation module may analyse the received Knowledge Graphs and their corresponding weight matrices together with confidence scores for the respective actions [determining, based on the macro feature, whether the target subgraph satisfies a requirement, and upon determining that the target subgraph does not satisfy the requirement] and, based on the differences between the weight matrices and the confidence scores, determine which of the actions is most effective to turn a Knowledge Graph into the state of interest [modifying the target subgraph or re-obtaining a target subgraph from the shared knowledge graph].)
Additionally, Kim teaches wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain, (in [0027] A knowledge graph (KG) is a collection of machine-readable descriptions of interlinked entities including, for example, real-world objects, events, situations, or concepts. Many AI-based applications, such as the example illustrated in FIG. 1, rely on knowledge graphs to provide background knowledge and concept & entity awareness to enable a more accurate interpretation of text and speech data (such as the natural language inquiry 119 in the example of FIG. 1). Knowledge graphs can be expressed using graphs with nodes connected by labeled and directed edges [wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain]. Each node expresses an entity and each labeled & directed edge represents a relationship among the entities [wherein selecting several nodes and their edges from the shared knowledge graph based on one or more entity types involved in the target service domain].)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the present application to combine the teachings of Kim and Tyler for the same reasons disclosed above.
Regarding claim 6, the rejection of claim 1 is incorporated and Tyler in combination with Kim teaches the method of method according to claim 1, wherein the target data processing task is entity classification, inter-entity relationship prediction, or entity set mining. (in [0044] The graphs that were analyzed in the past (i.e. G={(KG.sub.t.sup.t1, KG.sub.t+x.sup.t1) . . . (KG.sub.t.sup.tn, KG.sub.t+x.sup.tn)}, as introduced above, with the corresponding weight matrices, and classification results [wherein the target data processing task is entity classification]) can be used as training data… [0050] All of these KGs are analyzed by the GA module 110, i.e., the corresponding weight matrix w and the classification results m are computed [wherein the target data processing task is entity classification], as described already above…. Hence, by comparing the corresponding matrices it is possible to identify the action which influences the future the most in respect of the interest of the user 102… [0053] Further, according to an embodiment, the evaluation module 114 may also compare the KGs across the pairs and compute the respective weight matrix. Based on the differences between the weight matrices and the confidence scores, the evaluation module 114 computes which action is most effective to turn the KG into the State of Interest…)
Regarding independent claims 13 and 14, Tyler teaches a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to: … (in [0014] ... Furthermore, the system may comprise computer-implemented tools that are configured to analyze the information and to predict future events and developments.)
Additionally Kim teaches a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to: … (in [0025] FIG. 1 illustrates an example of a system for updating and modifying knowledge graphs & machine-learning models and for using knowledge graphs & machine-learning models. A computer-based system with a controller 101 includes an electronic processor 103 and a non-transitory computer-readable memory 105 [a system, comprising a memory and a processor, wherein the memory stores executable instructions that, in response to execution by the processor, cause the processor to: … and a non-transitory computer-readable storage medium, comprising instructions stored therein that, when executed by a processor of a computing device, cause the processor to]. The controller 101 is communicatively coupled to one or more user interface devices 107 including, for example, a display screen, a keyboard, a mouse, and/or a touch-sensitive display. The controller 101 is also communicatively coupled to a communication interface device 109 to facilitate communication between the controller 101 and other computer-based systems and networks. In the example of FIG. 1, the controller 101 is configured to communicate with one or more cloud computing systems 111 to access knowledge graphs 113 and/or machine-learning models 115 stored on the cloud computing systems 111…)
The remaining limitations of claims 13 and 14 are similar to claim 1 limitations and are rejected under the same rationale.
Claim 5 is rejected under 35 U.S.C. 103 as being unpatentable over Sztyler et al. (US 20240095805, hereinafter ‘Tyler’) in view of Kim (US 20220284309) in further view of Liu et al. (NPL: Personalized Navigation and Random Walk on a Complex Heterogeneous Graph, hereinafter ‘Liu’).
Regarding claim 5, the rejection of claim 1 is incorporated and Tyler in combination with Kim teaches the method of method according to claim 1, wherein the target subgraph is a heterogeneous graph; (in [0040] Everything which is logged/observed (e.g. through a sensor network) by the AI unit 104 is transformed into a Knowledge Graph (KG), i.e., a set of triples. In this regard, the AI unit 104 may be seen as an information extraction pipeline that (continuously or periodically) extracts new triples. This set of triples represents entities and objects, relation between them, and corresponding attributes. According to some embodiments, the AI unit 104 may be configured to store the information found in a Temporal Knowledge Graph (TKG), i.e. a KG that also contains temporal facts indicating relationships among entities and objects at different times. Embodiments of the invention combine (i.e. take into account) the time-dependent triples with a list of actions [wherein the target subgraph is a heterogeneous graph combined from a list of action as heterogenous graph containing different relations amount different entries and objects at different times] and a scope of desired outcomes [to obtain a target subgraph] (i.e. States of Interest) to compute recommendations, as will be described in more detail below…)
and processing the target subgraph to extract one or more graph features comprises: splitting the target subgraph into a plurality of homogeneous graphs; and separately processing the homogeneous graphs to extract one or more graph features. (in [0052] According to the illustrated embodiment, the output of the DA module 112 is forwarded to the evaluation module 114. The evaluation module 114, which may be implemented in form of a neural network, is configured to analyze the three (or more, depending on the number of time steps applied by the DA module 112) pairs of KGs and the corresponding weight matrices together with the confidence scores [and processing the target subgraph to extract one or more graph features comprises: splitting the target subgraph into a plurality of homogeneous graphs as the split of pairs of KGs separated to be analyzed and processed by the neural network that can be scored and weighted for respective actions] for the respective actions… [0054] The result, i.e. the output of the evaluation module 114, is a ranked list of actions, the corresponding effect (i.e. how the KG changes), and the outcome (i.e. the State of Interest) [separately processing the homogeneous graphs to extract one or more graph features as the ranked listed actions associated with the graph features with the target outcome to reliably make recommendations]. By using a Temporal Knowledge Graph (TKG) based concept to model the (time-dependent) entities, relations, and attributes, the computed recommendations are highly reliable.)
Liu expressly teaches the graph transfer process including the transfer of an heterogenous graph into a homogeneous graph, in Abstract: … In this study, we propose a novel approach for a personalized randomwalk over a complex heterogeneous graph; which we refer to as Personalized Graph Navigation (PGN). Unlike earlier expert-guided random walk approaches, by using an EM framework, PGN esti mates the personalized usefulness probability distribution for each edge type (the latent variable), which can be used as a random walk navigation profile for a user on a heterogeneous graph. While PGN can cope with information retrieval/recommendation problems at a low cost, this method also transforms a complex heterogeneous graph into a homogeneous graph [processing the target subgraph to extract one or more graph features comprises: splitting the target subgraph into a plurality of homogeneous graphs; and separately processing the homogeneous graphs to extract one or more graph features],… pg. 218: Left Col. : … In this study, we propose a novel method, Personalized Graph Navigation (PGN), for personalized random walk, search and recommendation over a complex heterogeneous graph. Given a user and a specific search or recommendation task, we simplify the problem by introducing a new latent variable, the personalized edge type usefulness probability distribution, which enables an innovative approach to estimating a user profile. More importantly, this latent variable allows a complex heterogeneous graph to be simplified to a homogeneous graph where the random walk probability is navigated using a classical edge transitioning probability combined with a novel personalized edge type usefulness probability…Third, by using the new latent variable (personalized edge type usefulness probability distribution), a complex heterogeneous graph can be converted to a homogeneous graph for random walk and ranking… Graph mining has proven to be one of the most efficient means for information search and recommendation and a number of algorithms are well documented... Meng et al. [22], applied the personalized random walk to a multi-layer graph for citation recommendation, where different kinds information are treated equally in the graph. However, most existing work on ranking is based on a homogeneous graph containing a single type of object or link [processing the target subgraph to extract one or more graph features comprises: splitting the target subgraph into a plurality of homogeneous graphs; and separately processing the homogeneous graphs to extract one or more graph features]. Intuitively, with multiple kinds of data available, a heterogeneous graph may be a better data structure for search and recommendation tasks…
Liu, Kim and Tyler are analogous art because both involve developing information retrieval and processing techniques using machine learning systems and algorithms.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of the prior art of developing information retrieval and processing techniques implementing knowledge graphs for machine learning algorithms, as disclosed by Liu with the method of developing information retrieval and processing techniques for machine learning task, as collectively disclosed by Kim and Tyler.
One of ordinary skill in the arts would have been motivated to combine the methods disclosed by Liu, Kim and Tyler as noted above. Doing so allowing for developing allow complex heterogeneous graph to be simplified to a homogeneous graph where the random walk probability is navigated using a classical edge transitioning probability combined with a novel personalized edge type usefulness probability, (Liu, Abstract and pg. 218 Left Col.).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Murali et al. (US 20230216967): teaches in [0037] It should be noted that in one example, a prediction model may take an embedding/vector as input(s). In one example, the prediction model may also take the graph itself, a sub-graph thereof, or one or more features extracted directly from the graph or derived from the graph, as an additional input or inputs. In such case, a “set of input factors” may include an embedding/vector and the graph (or sub-graph), or other features extracted/derived therefrom. In addition, labels may be added to at least a portion of a plurality of sets of input factors (e.g., labels of “fraud” or “no fraud”) as stored in DB(s) 136. In one example, the graph embedding process is learned/trained as part of the prediction model (e.g., where the extracted graph features and/or dimensionality of the vector(s)/embedding(s) are optimized for the specific task of the prediction model) or are trained as a separate process from the prediction model (e.g., guided graph embedding, where dimensionality of vector(s) and/or other hyperparameters is/are provided). Alternatively, dimensionality may be based upon selected loss criteria (e.g., an allowable level/percentage of loss of information).
Lee et al. (US 11544535):teaches in 5:51-58: As used herein, the term “graph” refers to a structure that models pairwise relations among entities in a dataset. A graph includes a set of nodes (also referred to as vertices or points) and a set of undirected or directed edges (also referred to as arcs or lines) connecting the set of nodes. When the edges are directed, the graph is a directed graph. When the edges are undirected, the graph is an undirected graph. Each node in a graph corresponds to an entity in the dataset represented by the graph, and features of a node correspond to attributes of the corresponding entity. A dataset representable by a graph is referred to as a graph-structured dataset… 6:12-45: As used herein, the term “feature matrix” refers to a matrix that describes a set of features or attributes of a set of entities (e.g., users or other objects) in a graph-structured dataset. Each entity is represented by a node (or a vertex, which is used interchangeably with the term “node” in this disclosure) in a graph. For example, each column of a feature matrix corresponds to a feature, and each row of the feature matrix is an entry that represents one entity, where each element in the row describes a corresponding feature or attribute of the entity, such as an identification or a characteristic of the entity… As used herein, the term “neighborhood” refers to a subgraph adjacent to a node in a graph. In some examples, a Kth-order neighborhood of a node refers to the set of nodes that lie within a distance K (i.e., K hops or K steps) from the node…
Any inquiry concerning this communication or earlier communications from the examiner should be directed to OLUWATOSIN ALABI whose telephone number is (571)272-0516. The examiner can normally be reached Monday-Friday, 8:00am-5:00pm 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, Michael Huntley can be reached at (303) 297-4307. 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.
/OLUWATOSIN ALABI/Primary Examiner, Art Unit 2129
Prosecution Insights