# US 12,001,488 B2 · Efficient data storage and retrieval system > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,001,488 B2 - Document ID: US12001488B2 - Status: Granted - Kind: B2 - Application number: 18/048,727 - Filed: 2022-10-21 - Published: 2024-04-25 - Granted: 2024-06-04 - Earliest claimed: 2022-10-21 - Assignees: Accenture Global Solutions Limited - Applicants: Accenture Global Solutions Limited - Inventors: Tianyou Zhang; Ramine Tinati; Nisith Singh; Ming Yang Tham; Hai Le; Sidharth Mittal - Web page: https://patentgrep.com/patent/US12001488B2 ## CPC - G06F 16/367: {Ontology} - G06F 16/906: Clustering; Classification - G06F 16/90344: {by using string matching techniques} - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract In some implementations, a device may receive data associated with a set of taxonomies, wherein a taxonomy, of the set of taxonomies, represents a classification of a set of relationships of data entries of the data. The device may integrate the set of taxonomies to generate a meta-taxonomy of the data. The device may generate a data graph of the data based on integrating the set of taxonomies to generate the meta-taxonomy, wherein the data graph is based on a graph-based search model that is associated with at least one of: taxonomy-based filtering, metadata attribute prioritization, or semantic matching. The device may store the data graph. ## Claims ### Claim 1 A device, comprising: one or more memories; and one or more processors, communicatively coupled to the one or more memories, configured to: receive data associated with a set of taxonomies, wherein a taxonomy, of the set of taxonomies, represents a classification of a set of relationships of data entries of the data; integrate, using a taxonomy harmonization model, the set of taxonomies to generate a meta-taxonomy of the data, wherein the meta-taxonomy includes one or more relationships between a first classification of a first taxonomy, of the set of taxonomies, and a second classification of a second taxonomy, of the set of taxonomies, wherein the first classification and the second classification are linked with the one or more relationships based on a similarity between the first classification and the second classification; receive a search query; search, by the device and using the meta-taxonomy, the data to identify one or more data entries, of the data, matching the search query; and transmit, by the device, a response identifying at least one data entry of the one or more data entries matching the search query. ### Claim 2 (depends on 1) The device of claim 1, wherein the one or more processors, when searching the data, are configured to: perform taxonomy filtering, using the meta-taxonomy, to identify a subset of the data to search; and search the subset of the data to identify the one or more data entries. ### Claim 3 (depends on 1) The device of claim 1, wherein the one or more processors, when searching the data, are configured to: search a set of contextual tags to identify the one or more data entries. ### Claim 4 (depends on 1) The device of claim 1, wherein the one or more processors, when integrating the set of taxonomies, are configured to: integrate the set of taxonomies using a Jaccard similarity technique to identify overlapping taxonomical tagging; and generate the meta-taxonomy with the one or more relationships based on the overlapping taxonomical tagging. ### Claim 5 (depends on 1) The device of claim 1, wherein the one or more processors, when integrating the set of taxonomies, are configured to: integrate the set of taxonomies using an overlap similarity technique to identify an overlapping text-match between a first term of the first classification and a second term of the second classification; and generate the meta-taxonomy with the one or more relationships based on the overlapping text-match. ### Claim 6 (depends on 1) The device of claim 1, wherein the one or more processors, when integrating the set of taxonomies, are configured to: integrate the set of taxonomies using an overlap similarity technique to identify overlapping without a text-match between a first term of the first classification and a second term of the second classification, such that the second classification is at least partially a subset of the first classification; and generate the meta-taxonomy with the one or more relationships based on the overlapping without the text-match. ### Claim 7 (depends on 1) The device of claim 1, wherein the one or more processors, when integrating the set of taxonomies, are configured to: integrate the set of taxonomies using a semantic similarity technique to identify a term-match between a first semantic meaning of a first term of the first classification and a second semantic meaning of a second term of the second classification; and generate the meta-taxonomy with the one or more relationships based on the term-match. ### Claim 8 A method, comprising: receiving, by a device, data associated with a set of taxonomies, wherein a taxonomy, of the set of taxonomies, represents a classification of a set of relationships of data entries of the data; integrating, by the device and using a taxonomy harmonization model, the set of taxonomies to generate a meta-taxonomy of the data, wherein the meta-taxonomy includes one or more relationships between a first classification of a first taxonomy, of the set of taxonomies, and a second classification of a second taxonomy, of the set of taxonomies, wherein the first classification and the second classification are linked with the one or more relationships based on a similarity between the first classification and the second classification; generating, by the device, a data graph of the data based on integrating the set of taxonomies to generate the meta-taxonomy, wherein the data graph is based on a graph-based search model that is associated with at least one of: taxonomy-based filtering, metadata attribute prioritization, or semantic matching; and storing the data graph. ### Claim 9 (depends on 8) The method of claim 8, further comprising: searching the meta-taxonomy to detect a particular taxonomy relating to a search query; searching the data graph of the meta-taxonomy for one or more taxonomy nodes associated with the particular taxonomy; and identifying one or more data entries matching the search query based on the one or more taxonomy nodes associated with the particular taxonomy; and providing an output based on the one or more data entries. ### Claim 10 (depends on 8) The method of claim 8, wherein searching the meta-taxonomy comprises: applying a natural language processing model to the data graph to prioritize a set of data entries identified from the meta-taxonomy as potential matches for a search query; and identifying one or more data entries matching a search query based on the prioritization of the set of data entries. ### Claim 11 (depends on 8) The method of claim 8, further comprising: searching the meta-taxonomy to identify one or more data entries matching a search query; and outputting the one or more data entries in an order based on at least one of: a semantic meaning technique, an entity boosting technique, a popularity search technique, or a graph-based filtering technique. ### Claim 12 (depends on 8) The method of claim 8, further comprising: searching the meta-taxonomy to identify a search intent based on a search query; generating a set of attributes based on the search intent; and identifying one or more data entries matching the search query based on the set of attributes; and providing an output based on the one or more data entries. ### Claim 13 (depends on 8) The method of claim 8, further comprising: searching the meta-taxonomy using a vector search technique applied to the data graph; and providing an output based on searching the meta-taxonomy using the vector search technique. ### Claim 14 (depends on 8) The method of claim 8, wherein further comprising: determining whether a search query satisfies a natural language processing feasibility criterion; and applying a natural language processing searching technique or a full text searching technique based on whether the search query satisfies the natural language processing feasibility criterion. ### Claim 15 (depends on 14) The method of claim 14, wherein the natural language processing feasibility criterion is based on at least one of: in which language the search query is provided or a quantity of nouns in the search query. ### Claim 16 (depends on 14) The method of claim 14, wherein the natural language processing searching technique includes at least one of: intent-driven dedicated search, named entity recognition boosting, or graph-based popularity searching. ### Claim 17 (depends on 14) The method of claim 14, wherein the full text searching technique includes at least one of full text searching or graph-based popularity searching. ### Claim 18 A non-transitory computer-readable medium storing a set of instructions, the set of instructions comprising: one or more instructions that, when executed by one or more processors of a device, cause the device to: receive data associated with a set of taxonomies, wherein a taxonomy, of the set of taxonomies, represents a classification of a set of relationships of data entries of the data; integrate, using a taxonomy harmonization model, the set of taxonomies to generate a meta-taxonomy of the data, wherein the meta-taxonomy includes one or more relationships between a first classification of a first taxonomy, of the set of taxonomies, and a second classification of a second taxonomy, of the set of taxonomies, wherein the first classification and the second classification are linked with the one or more relationships based on a similarity between the first classification and the second classification; receive a search query; identify a user profile for a user associated with the search query; and search the meta-taxonomy to identify one or more data entries, of the data, matching the search query based on the user profile; and transmit a response identifying at least one data entry of the one or more data entries matching the search query. ### Claim 19 (depends on 18) The non-transitory computer-readable medium of claim 18, wherein the one or more instructions, that cause the device to identify the user profile, cause the device to: receive information identifying a user preference; identify a set of other user profiles associated with other users based on the user preference; and import one or more other user preferences of the set of other user profiles into the user profile for the user. ### Claim 20 (depends on 18) The non-transitory computer-readable medium of claim 18, wherein the one or more instructions, that cause the device to identify the user profile, cause the device to: identify a knowledge graph representing a set of historical transactions associated with the user, wherein the knowledge graph includes a set of weightings relating to at least one of a recency, a frequency, or a rating of a historical transaction of the set of historical transactions.