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 .
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 3/23/2026 has been entered.
Status of Claims
Claims 1-20 are pending of which claims 1, 11 and 20 are in independent form.
Claims 1-20 are rejected under 35 U.S.C. 101.
Claims 1-20 are rejected under 35 U.S.C. 103.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Regarding the 35 USC 101 (Abstract Idea), remarks made by the applicant.
Examiner specifies that, the newly added amendments/arguments do not overcome the 35 USC 101 rejection.
With respect to step 2A, prong one (Judicial Exception), the claims recite an abstract idea, law of nature, or natural phenomenon. Specifically, the following limitations recite mathematical concepts and/or mental processes and/or certain methods of organizing human activity.
The claim recites sequence of operations that amount to information organization, searching, evaluation, and ranking directed to an abstract idea:
Detecting a signal indicating tenant eligibility;
Identifying tenant specific content based on criteria;
Generating semantic vectors (vectorization);
Storing and replicating vectors between storage locations;
Determine readiness based on comparison; and
Enabling querying after the readiness is met.
There steps amount to:
Collecting and organizing data
Vectorization of content
Copying and managing data across storage location
Comparing current and expected data
Determining whether sufficient data exists (readiness)
Allowing access (based on readiness)
These steps collectively fall into recognized abstract idea:
Mental process: determine eligibility; selecting content based on criteria; evaluating a completeness metric against a threshold.
Mathematical concept/algorithm: determining whether index is sufficient, comparing actual vs expected, decide whether its ready to use.
Mathematical Algorithm: semantic vectors, embeddings
Information Analysis and Evaluation: determine readiness based on comparison
Data Collecting and Organization: identifying content, storing, replicating, and organizing into index.
These operations are performed are directed to managing the lifecycle of indexed information on generic off the shelf technology.
There are no steps performed that provides a technical improvement to the computing system itself (improved data structure, improved network architecture, improved hardware performance, improved streaming protocol).
Thus, the claims recite an abstract idea (mental process/mathematical algorithm/information analysis/data collection).
With respect to step 2A, Prong Two (Particular Application), the claims do not recite additional elements that integrate the judicial exception into a practical application. The following limitations are considered “additional elements” and explanation will be given as to why these “additional elements” do not integrate the judicial exception into a practical application.
The claims recite the use of:
A “cloud based” environment;
primary and secondary index storage;
replication for distributed querying;
enabling queries based on the threshold.
These components merely use conventional computer components as tools to execute the abstract idea:
these are generic computing environments and storage components;
data replication and distributed querying are conventional cloud computing practices;
the claims do not improve: vector generation algorithms; index structures; network replication protocols; storage efficiency or latency in a technical way.
These do not provide a technical improvement; they are recited at a high level with no specificity (e.g. no improved: vector generation algorithms; index structures; network replication protocols; storage efficiency or latency in a technical way)
The limitations fail to transform the exception into a practical application. There is also no improvements to computer functionality or any specific technical solution to a computer centric problems.
Instead the claims recite conventional and generic computer functions performed in a routine manner, which does not amount to a practical application.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
The claim(s) recite(s) management of semantic indexes in a distributed cloud computing environment for enhancing search and query functionalities for tenant- specific data within a scalable vector database framework.
With respect to step 1 of the patent subject matter eligibility analysis, the claims are directed to a process, machine, manufacture, or composition of matter.
Independent claim 1 is directed to a method, which is a process.
Independent Claim 11 is directed to a computing apparatus, including a processor; and a memory, which is a machine, and directed to one of the 4 categories of patent eligible subject matter.
Independent claim 20 is directed to a non-transitory computer-readable storage medium, which is a process, which is directed to one of the 4 categories of patent eligible subject matter.
All other claims depend on claims 1, 11 and 20. As such, claims 1-20 are directed to a statutory category.
Regarding claims 1, 11 and 20:
With respect to step 2A, prong one (Judicial Exception), the claims recite an abstract idea, law of nature, or natural phenomenon. Specifically, the following limitations recite mathematical concepts and/or mental processes and/or certain methods of organizing human activity.
The claim recites sequence of operations that amount to information organization, searching, evaluation, and ranking directed to an abstract idea:
Detecting a signal indicating tenant eligibility;
Identifying tenant specific content based on criteria;
Generating semantic vectors (vectorization);
Storing and replicating vectors between storage locations;
Determine readiness based on comparison; and
Enabling querying after the readiness is met.
There steps amount to:
Collecting and organizing data
Vectorization of content
Copying and managing data across storage location
Comparing current and expected data
Determining whether sufficient data exists (readiness)
Allowing access (based on readiness)
These steps collectively fall into recognized abstract idea:
Mental process: determine eligibility; selecting content based on criteria; evaluating a completeness metric against a threshold.
Mathematical concept/algorithm: determining whether index is sufficient, comparing actual vs expected, decide whether its ready to use.
Mathematical Algorithm: semantic vectors, embeddings
Information Analysis and Evaluation: determine readiness based on comparison
Data Collecting and Organization: identifying content, storing, replicating, and organizing into index.
These operations are performed are directed to managing the lifecycle of indexed information on generic off the shelf technology.
There are no steps performed that provides a technical improvement to the computing system itself (improved data structure, improved network architecture, improved hardware performance, improved streaming protocol).
Thus, the claims recite an abstract idea (mental process/information organization).
With respect to step 2A, Prong Two (Particular Application), the claims do not recite additional elements that integrate the judicial exception into a practical application. The following limitations are considered “additional elements” and explanation will be given as to why these “additional elements” do not integrate the judicial exception into a practical application.
The claims recite the use of:
A “cloud based” environment;
primary and secondary index storage;
replication for distributed querying;
enabling queries based on the threshold.
These components merely use conventional computer components as tools to execute the abstract idea:
these are generic computing environments and storage components;
data replication and distributed querying are conventional cloud computing practices;
the claims do not improve: vector generation algorithms; index structures; network replication protocols; storage efficiency or latency in a technical way.
These do not provide a technical improvement; they are recited at a high level with no specificity (e.g. no improved: vector generation algorithms; index structures; network replication protocols; storage efficiency or latency in a technical way)
The limitations fail to transform the exception into a practical application. There is also no improvements to computer functionality or any specific technical solution to a computer centric problems.
Instead the claims recite conventional and generic computer functions performed in a routine manner, which does not amount to a practical application.
With respect to Step 2B. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. The recited components are merely generic computer/database elements performing their routine, well-understood, and conventional functions. See Alive, MPEP 2016.05(d).
The steps mentioned in the independent claims merely constitutes generic indexing process. However, claims fail to provide:
specialized hardware;
unconventional data structure;
technical improvement to indexing and/or querying;
non-routine replication mechanism;
specific algorithm enhancement.
Courts have consistently helped such high level information management operations are conventional.
The claims recite only, without significantly more, listing specific signals and external signals are result oriented and functional (without technical implementation), no practical algorithm, signal generation technique, or ranking mechanism. All are routine, conventional operations business/ market place logic.
Considering claims as a whole, the ordered combination of elements also reflects nothing more than the typical workflow of distributed systems, and therefore DOES NOT add “significantly more” than the abstract idea.
Such generic, high‐level, and nominal involvement of a computer or computer‐based elements for carrying out the invention merely serves to tie the abstract idea to a particular technological environment, which is not enough to render the claims patent‐eligible, as noted at pg.74624 of Federal Register/Vol. 79, No. 241, citing Alice, which in turn cites Mayo. Further, See, e.g., Alice Corp. Pty. Ltd. v. CLS Bank Int'l, 134 S. Ct. 2347, 2359‐60, 110 USPQ2d 1976, 1984 (2014). See also OIP Techs. v. Amazon.com, 788 F.3d 1359, 1364, 115 USPQ2d 1090, 1093‐94 (Fed. Cir. 2015) ("Just as Diehr could not save the claims in Alice, which were directed to 'implement[ing] the abstract idea of intermediated settlement on a generic computer', it cannot save O/P's claims directed to implementing the abstract idea of price optimization on a generic computer.") (citations omitted). See also, Affinity Labs of Texas LLC v. DirecTV LLC, 838 F.3d 1253, 1257‐1258 (Fed. Cir. 2016) (mere recitation of a GUI does not make a claimpatent‐eligible); Intellectual Ventures I LLC v. Capital One Bank, 792 F.3d 1363, 1370 (Fed. Cir. 2015) ("the interactive interface limitation is a generic computer element".).
The additional elements are broadly applied to the abstract idea at a high level of generality ("similar to how the recitation of the computer in the claims in Alice amounted to mere instructions to apply the abstract idea of intermediated settlement on a generic computer,") as explained in MPEP § 2106.05(f)) and they operate in a well‐understood, routine, and conventional manner.
MPEP § 2106.0S(d)(II) sets forth the following:
The courts have recognized the following computer functions as well-understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity.
• Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec ... ; TLI Communications LLC v. AV Auto. LLC ... ; OIP Techs., Inc., v. Amazon.com, Inc ... ; buySAFE, Inc. v. Google, Inc ... ;
• Performing repetitive calculations, Flook ... ; Bancorp Services v. Sun Life ... ;
• Electronic recordkeeping, Alice Corp ... ; Ultramercial ... ;
• Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc ... ;
• Electronically scanning or extracting data from a physical document, Content Extraction and Transmission, LLC v. Wells Fargo Bank ... ; and
• A web browser's back and forward button functionality, Internet Patent
• Corp. v. Active Network, Inc. ...
. . . Courts have held computer-implemented processes not to be significantly more than an abstract idea (and thus ineligible) where the claim as a whole amounts to nothing more than generic computer functions merely used to implement an abstract idea, such as an idea that could be done by a human analog (i.e., by hand or by merely thinking).
In addition, when taken as an ordered combination, the ordered combination adds nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements integrate the abstract idea into a practical application. Their collective functions merely provide conventional computer implementation. Therefore, when viewed as a whole, these additional claim elements do not provide meaningful limitations to transform the abstract idea into a practical application of the abstract idea or that the ordered combination amounts to significantly more than the abstract idea itself.
The dependent claims have been fully considered as well, however, similar to the findings for claims above, these claims are similarly directed to the “Mental Processes” grouping of abstract ideas set forth in the 2019 PEG, without integrating it into a practical application and with, at most, a general purpose computer that serves to tie the idea to a particular technological environment, which does not add significantly more to the claims. The ordered combination of elements in the dependent claims (including the limitations inherited from the parent claim(s)) add nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide conventional computer implementation. Accordingly, the subject matter encompassed by the dependent claims fails to amount to significantly more than the abstract idea.
Looking at the claim as a whole does not change this conclusion and the claim is ineligible.
Examiner further indicates that the remaining claims also fall with 35 USC 101 (abstract idea) for at least the similar reasons.
All the dependent claims perform generic functions:
Claims 2 and 12: (Bootstrap/Initialization of Index)
falls under:
Data collection and organization
Pre-processing logic
Claims 3 and 13: (Content Type Selection for Vectorization)
falls under:
data selection
Information categorization
Claims 4 and 14 (Use of Vector Database/Graph Structure):
Falls under:
Data storage and organization
Generic computer implementation
Claims 5 and 15: (Incremental Updates/Change Detection):
Falls under:
Data maintenance
Monitoring and updating information
Claims 6 and 16: (Deprovisioning/Deletion):
Falls under:
Data management
Conditional deletion logic
Claim 7, 8, 17 and 18: (Completeness/Integrity-Based Query Enablement):
Falls under:
Mental Process (evaluation/threshold decision)
Mathematical Algorithm (percentage/metrics)
Data Analysis
Claims 9 and 19: (Primary and Secondary Storage Roles):
Falls under:
Generic System Architecture
Data routing and organization
Claim 10: (Monitoring and Tuning Based on Usage):
Falls under:
Mental Process (analysis and decision making)
Data analysis and optimization
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.
Claim(s) 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Crabtree; Jason et al. (US 20240386015 A1) [Crabtree] in view of TEOFILI; TOMMASO et al. (US 20180095989 A1) [Teofili] in view of Liu; Zhi (US 20250005896 A1) [Liu] in view of LI; Mingqin et al. (US 20210406321 A1) [Li].
Regarding claims 1, 11 and 20, Crabtree discloses, a method for managing a lifecycle of a semantic index for tenants in a cloud-based environment, the method comprising: detecting a signal indicating a tenant eligibility for semantic indexing (reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies ¶ [0030]-[0032]);
in response to detecting the signal, identifying tenant-specific content for vectorization based on criteria (reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies; processing user queries and returning relevant results by leveraging the vector semantic indices, knowledge graphs, and contextual information ¶ [0030]-[0032]. Artificial intelligence techniques (e.g., large language models) to create, update, and align and evolve ontologies and curate ontological data from diverse data sources while also creating vector semantic indices and traditional database indices ¶ [0090]-[0092]);
generating semantic vectors from the identified tenant-specific content (reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies; processing user queries and returning relevant results by leveraging the vector semantic indices, knowledge graphs, and contextual information ¶ [0030]-[0032]. Artificial intelligence techniques (e.g., large language models) to create, update, and align and evolve ontologies and curate ontological data from diverse data sources while also creating vector semantic indices and traditional database indices ¶ [0090]-[0092]) and storing the semantic vectors in a primary index storage associated with tenant (Embeddings are dense vector representations that capture the semantic meaning and relationships of data points. Vector databases 2128 store and index these embeddings for efficient retrieval and similarity search ¶ [0144]. The vector database 315 is responsible for efficiently storing, comparing, and retrieving a large plurality of embeddings (i.e., vectors). Vector database 315 may be any suitable vector database system known to those with skill in the art including, but not limited to, open source systems like Pinecone, Weaviate, Vespa, and Qdrant. According to the embodiment, embedding model 315 may also receive a user query from experience curation 340 and vectorize it where it may be stored in vector database 320 ¶ [0253]-[0254]. Also see ¶ [0271], [0304]);
configured for distributed querying and physically or logically separated from the primary index storage (replication techniques to distribute the index across multiple nodes, improving scalability… This may comprise developing query parsing and understanding components to interpret user queries and match them against the indexed ontology concepts and relationships, and the implementation of query expansion techniques, such as query rewriting and query suggestion, to improve the recall and relevance of search results by considering synonyms, related concepts, and user intent ¶ [0307]. The vectorized data may flow from the embedding node 930 to a data storage node 950. Data storage node 950 may select the appropriate vector database 980 in which to store the vectorized context data. An input node 940 may allow for a user to submit a query to the workflow. The user query can be sent to data embedding node 930 where it may be vectorized and sent to data storage node 950 for storage in the vector database. The user query can also be sent to a model node 960 which contains the selected model(s) which will process the user query along with any relevant context data obtained from data storage node vector database 980 ¶ [0271]);
building a semantic index from the semantic vectors stored in the secondary index storage (reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies; processing user queries and returning relevant results by leveraging the vector semantic indices, knowledge graphs, and contextual information ¶ [0030]-[0033]. Artificial intelligence techniques (e.g., large language models) to create, update, and align and evolve ontologies and curate ontological data from diverse data sources while also creating vector semantic indices and traditional database indices ¶ [0090]-[0092]).
However, Crabtree does not explicitly facilitate replicating the semantic vectors from the primary index storage to a secondary index storage.
Teofili discloses, replicating the semantic vectors from the primary index storage to a secondary index storage (By treating file storage specifications (or at least a portion thereof) as “words” in the language model, replication vectors can be determined based on the file storage specifications. Instead of determining the relationship of the file storage specifications based on ordering within a document, the relationship can be based on proximity of the replication requests in a replication session. When a replication request is received from a user, the replication vectors can be used to determine a semantic similarity between the received replication request and one or more additional replication requests [Abstract]. Also see ¶ [0003] and [0027]-[0032]).
It would have been obvious to one ordinary skilled in the art at the time of the filing of the present invention to combine the teachings of the cited references because Teofili's system would have allowed Crabtree to facilitate replicating the semantic vectors from the primary index storage to a secondary index storage. The motivation to combine is apparent in the Crabtree's reference, because there is a need to improve replicating documents using less intensive processes.
However, Crabtree nor Teofili explicitly facilitates maintaining the secondary index storage in [a non-queryable state until the semantic index satisfies a readiness threshold]; computing, by a query enablement module, an index [readiness state] based on a comparison between (i) a current quantity of semantic vectors confirmed as replicated to the secondary index storage, and (ii) an expected quantity of semantic vectors generated by the primary index storage.
Liu discloses, maintaining the secondary index storage [in a non-queryable state until the semantic index satisfies a readiness threshold] (maintaining a vector index by replicated and incremental updating index structures across nodes using reduplicated and reconstructed vector batches via batches ¶ [0065]-[0068]);
computing, by a query enablement module, an index [readiness state] based on a comparison between (i) a current quantity of semantic vectors confirmed as replicated to the secondary index storage, and (ii) an expected quantity of semantic vectors generated by the primary index storage (Liu: computing conditions by evaluating similarity between vectors and computing similarity scores. And comparing those scores to a threshold to determine whether vectors should be incorporated into the maintained index ¶ [0065], [0074]).
It would have been obvious to one ordinary skilled in the art at the time of the filing of the present invention to combine the teachings of the cited references because Liu's system would have allowed Crabtree and Teofili to facilitates maintaining the secondary index storage in [a non-queryable state until the semantic index satisfies a readiness threshold]; computing, by a query enablement module, an index [readiness state] based on a comparison between (i) a current quantity of semantic vectors confirmed as replicated to the secondary index storage, and (ii) an expected quantity of semantic vectors generated by the primary index storage. The motivation to combine is apparent in the Crabtree and Teofili's reference, because there is a strong need for improved methods and systems for indexing embedding vectors, particularly at above-billion scale for fast high-recall retrieval.
However, neither Crabtree, Teofili, or Liu explicitly facilitates in a non-queryable state until the semantic index satisfies a readiness threshold; readiness state based on; enabling execution of semantic queries on the secondary index storage only after the index readiness state satisfies a predetermined readiness criterion.
Li discloses, a non-queryable state until the semantic index satisfies a readiness threshold; readiness state based on (determining when index construction/merging is complete…designating the index is ready … upon determining the index is ready, loading it as a read index used for query processing ¶ [0040]);
enabling execution of semantic queries on the secondary index storage only after the index readiness state satisfies a predetermined readiness criterion (upon determining that the index is ready, the system loads the index as the read index for query execution, thereby enabling query processing using the constructed index ¶ [0040]).
It would have been obvious to one ordinary skilled in the art at the time of the filing of the present invention to combine the teachings of the cited references because Li's system would have allowed Crabtree, Teofili and Liu to facilitates a non-queryable state until the semantic index satisfies a readiness threshold; readiness state based on; enabling execution of semantic queries on the secondary index storage only after the index readiness state satisfies a predetermined readiness criterion. The motivation to combine is apparent in the Crabtree, Teofili and Liu's reference, because there is a need for a new methods and systems designed to utilize approximate nearest neighbor search algorithms and provide results available immediately within adding the content to a content repository
Regarding claims 2 and 12, the combination of Crabtree, Teofili, Liu, and Li discloses, in response to detecting the signal, initiating a bootstrap process for creating the semantic index based on data schema (Crabtree: reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies ¶ [0030]-[0032]) and metadata of the tenant-specific content (Crabtree: In an embodiment, the system uses resource description framework (RDF) or web ontology language (OWL) to define core semantics of the advertised products. It may further incorporate aspects to track and manage changes in the advertising content and metadata over time ¶ [0175]-[0178]. According to an embodiment, data aggregation and enrichment system 4220 may aggregate data from all sources into a centralized repository 4230 for real-time processing and analysis. System 4220 may also enrich data with contextual metadata such as timestamps, geolocation, user preferences, and activity history, to name a few ¶ [0438]).
Regarding claims 3 and 13, the combination of Crabtree, Teofili, Liu, and Li discloses, wherein identifying tenant-specific content includes selecting content types comprising documents, emails, chats, and images for vectorization (Crabtree: With respect to multimodal data fusion the process may perform data collection and preprocessing. For text input, the system can collect text data from user queries, chat interactions, emails, documents, etc., and preprocess the text input using NLP techniques like tokenization, stemming, and lemmatization. For sound input, the system can collect audio data from voice commands, phone calls, or recorded messages and preprocess the audio input by converting speech to text using Automatic Speech Recognition (ASR) and extract emotional tone using sentiment analysis. For imagery input, the system can collect image data from user uploads, camera feeds, or social media and preprocess images using computer vision techniques to identify objects, scenes, and facial expressions ¶ [0418], [0429]).
Regarding claims 4 and 14, the combination of Crabtree, Teofili, Liu, and Li discloses, utilizing a scalable vector database to store and query semantic embeddings of items in a graph structure (Crabtree: By bridging the gap between non-textual and textual data through labeling, the platform can take advantage of the rich semantic information captured by language models and embeddings into text or alternative media or domain formats. After labeling non-textual data (if applicable), the platform computes numerical embedding representations of the input data in a given format. These embeddings capture the semantic properties and relationships of the data in a dense vector format, enabling efficient storage, retrieval, and comparison. The computed embeddings may then be persisted in memory or in a database such as a vector database, which allows for fast and scalable similarity search (e.g., cosine, dot product, Euclidean, etc.) and other vector operations or graph operations or hybrid representations depending on the data type, representation, and elements such as facts, spatial or temporal dynamics of the systems and/or entities of interest ¶ [0142]. Embeddings are dense vector representations that capture the semantic meaning and relationships of data points. Vector databases 2128 store and index these embeddings for efficient retrieval and similarity search ¶ [0144]).
Regarding claims 5 and 15, the combination of Crabtree, Teofili, Liu, and Li discloses, detecting changes in the identified tenant-specific content by monitoring for additions, deletions, or modifications; vectorizing new or modified data to generate updated semantic vectors corresponding to the changes detected (Crabtree: Platform can update the vector embeddings generated by the neural network models by incorporating the attention-weighted knowledge representations ¶ [0353]. Automatically creating and updating ontologies by analyzing structured and unstructured data from multiple sources using natural language processing and machine learning or artificial intelligence techniques; reading, creating, and maintaining a vector semantic index of content elements linked to core symbolic concepts and relationships defined in the ontologies; processing user queries and returning relevant results by leveraging the vector semantic indices, knowledge graphs, and contextual information; storing and managing knowledge corpora that integrates information from ontologies, semantic indices, and external sources ¶ [0030]);
propagating the updated semantic vectors from the primary index storage to the secondary index storage (Teofili: When the above relationships are received by the neural network language mode, the replication vectors for the file storage specifications “/a/b”, “/a/d”, and “/e/c” can each be updated. Replication vectors are “word-vectors”, corresponding to the vector representation of a word within the neural network language model. After the update, the scalar product (such as a cosine similarity) of the replication vector corresponding to “/a/b” with the replication vector corresponding to “/a/d” can have a higher value than prior to when the above additional relationship was used for training the model ¶ [0027], [0034]);
applying updates to the semantic index using the updated semantic vectors, wherein the updates modify only portions of the semantic index affected by the changes; and deploying the updated semantic index for querying (Crabtree: According to an embodiment, automated index generator subsystem 3250 can integrate with the knowledge graph database 2129 to seamlessly update the index whenever new ontological information is added or modified, and leverage the model blending computing system 2125 to combine multiple indexing and ranking strategies, optimizing search performance across different domains and user preferences ¶ [0308]. The system employs natural language processing, machine learning, and artificial intelligence techniques (e.g., large language models) to create, update, and align and evolve ontologies and curate ontological data from diverse data sources while also creating vector semantic indices and traditional database indices ¶ [0090]).
Regarding claims 6 and 16, the combination of Crabtree, Teofili, Liu, and Li discloses, detecting a tenant deprovisioning corresponding to the semantic index; and in response to detecting the tenant deprovisioning, deleting the semantic index in the primary index storage and the secondary index storage (Li: Multiple replicas of an approximate nearest neighbor index may be provided in order to provide a high service of availability. For example, one replica may be primary, and the others are secondary. During a vector update (add/delete) scenario, the vector may be copied from primary to secondary replicas. If data in an ANN index is lost, for example resulting from a hardware crash, one replica, ANN index file copy may support a fast recovery between vector index replicas ¶ [0032]. Also see ¶ [0034]-[0036]).
Regarding claims 7 and 17, the combination of Crabtree, Teofili, Liu, and Li discloses, wherein enabling the semantic queries further comprises: tracking a completeness or integrity of the semantic index to determine when the index is ready for serving the semantic queries; and activating a semantic query functionality based on the completeness or integrity of the semantic index exceeding a predetermined index completeness threshold (Crabtree: According to the aspect, an ontology quality assessment and refinement subsystem 3240 is present and configured to assess the quality of the generated ontologies. For example, the subsystem may apply ontology evaluation metrics such as consistency, completeness, and conciseness to assess the quality of the generated ontologies ¶ [0303]. Furthermore, the system may implement data fusion techniques, such as entity resolution, data deduplication, or schema matching, to merge and reconcile search results from multiple sources or iterations. The system can apply data quality assessment and cleansing techniques to ensure the consistency, accuracy, and completeness of the integrated search results before passing them to subsequent workflow stages ¶ [0335]).
Regarding claims 8 and 18, the combination of Crabtree, Teofili, Liu, and Li discloses, enabling the semantic queries further comprises: monitoring a completeness of the semantic index by calculating a completeness metric based on a percentage of expected data that is present within the semantic index; comparing the calculated completeness metric against a predefined threshold of completeness; and setting a query enablement flag to true when the calculated completeness metric meets or exceeds the predefined threshold of completeness, indicating that the semantic index has reached sufficient completeness to enable the semantic queries (Namiki: Further, the processing device 120 may have a function of calculating the degree of completion of the incomplete index 1121 and, in a case where the calculated degree of completion does not exceed a threshold, not processing the query by using the incompletion index 1121. Alternatively, the processing device 120 may be configured to calculate the degree of completion of the incomplete index 1121, process the query by using the incomplete index 1121 only in a case where the calculated degree of completion exceeds a threshold, and make the query inexecutable otherwise. As the degree of completion of the incomplete index 1121, the processing device 120 may use, for example, the number of the data 1111 already registered in the incomplete index 1121, or the ratio of the number of the data 1111 already registered in the incomplete index 1121 to the number of the data 1111 stored in the data part 111. Alternatively, in a case where the data part 111 is composed of a plurality of sub data parts, the processing device 120 may use, as the degree of completion of the incomplete index 1121, the number of sub data parts whose data are all registered in the incomplete index 1121 among the plurality of sub data parts, or the ratio of the number of sub data parts whose data are all registered in the incomplete index 1121 to the total number of the sub data parts ¶ [0041]. Also see ¶ [0054], [0120], [0146]).
Regarding claims 9 and 19, the combination of Crabtree, Teofili, Liu, and Li discloses, wherein the primary index storage is configured to ingest and process initial data to generate the semantic vectors, and the secondary index storage is configured to replicate and query the semantic vectors, the primary index storage serving as an initial repository for the semantic vectors and responsible for a vectorization process and initial index creation, (Teofili: By treating file storage specifications (or at least a portion thereof) as “words” in the language model, replication vectors can be determined based on the file storage specifications. Instead of determining the relationship of the file storage specifications based on ordering within a document, the relationship can be based on proximity of the replication requests in a replication session. When a replication request is received from a user, the replication vectors can be used to determine a semantic similarity between the received replication request and one or more additional replication requests [abstract]. Also see ¶ [0027]-[0032]),
the secondary index storage maintaining a copy of the semantic vectors from the primary index storage enabling distributed querying capabilities across the cloud-based environment (Crabtree: The cloud layer ensures data synchronization and replication mechanisms to keep the knowledge graph and ontologies consistent and up-to-date across the edge devices ¶ [0339], [0343]).
Regarding claim 10, the combination of Crabtree, Teofili, Liu, and Li discloses, continuously monitoring a performance of the semantic queries executed against the semantic index by collecting usage data and query response metrics; analyzing the collected usage data to identify patterns in the semantic index's performance and a relevance of query results; adjusting parameters of the semantic index based on the analyzing of the collected usage data to enhance an accuracy and efficiency of the semantic queries, wherein the adjustments include modifications to vectorization algorithms, index structure, or query processing methods; and implementing performance tuning measures that are responsive to the analyzing (Crabtree: This iterative refinement process allows the system to continuously learn and improve the accuracy and relevance of its links between vector semantic representations and ontological representations of data and to add to and curate multiple structured and even symbolic representations of data elements into effective knowledge corpora for specialized and broad-based search, reasoning and model training or utilization. The typical knowledge graph comprises nodes representing entities, concepts, and relationships, and edges representing the connections between them. The nodes are categorized into different types, such as classes, instances, and properties, based on their semantic roles. The edges are labeled with the specific relationships they represent, such as ‘is-a’, ‘part-of’, or ‘has-property’. This structured representation allows for efficient traversal and reasoning over the property graph. The system employs various reasoning and inference techniques, such as logical reasoning, rule-based inference, and graph pattern matching, to derive new knowledge and insights from the knowledge graph. For example, the system may use first-order logic to infer new facts based on existing relationships, or apply graph algorithms like shortest path or centrality measures to identify important entities and connections. Each of the nodes may also contain property information linking it to vectorized representation of its constituent data elements. Nodes and subgraphs may also be linked to supporting source content from which such elements were derived and may also reflect metadata about the provenance of the analysis (e.g., the model and its associated training data and author and history and license terms) which classified such elements or element properties into the symbolic knowledge base ¶ [0091]. Also see ¶ [0194], [0349], [0427]).
Conclusion
The examiner requests, in response to this Office action, support be shown for language added to any original claims on amendment and any new claims. That is, indicate support for newly added claim language by specifically pointing to page(s) and line no(s) in the specification and/or drawing figure(s). This will assist the examiner in prosecuting the application.
When responding to this office action, Applicant is advised to clearly point out the patentable novelty which he or she thinks the claims present, in view of the state of the art disclosed by the references cited or the objections made. He or she must also show how the amendments avoid such references or objections See 37 CFR 1.111(c).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOHAMMAD S ROSTAMI whose telephone number is (571)270-1980. The examiner can normally be reached Mon-Fri From 9 a.m. to 5 p.m..
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, Boris Gorney can be reached at (571)270-5626. 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.
4/4/2026
/MOHAMMAD S ROSTAMI/Primary Examiner, Art Unit 2154
Prosecution Insights