# US 12,216,664 B2 · Refining training sets and parsers for large and dynamic text environments > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,216,664 B2 - Document ID: US12216664B2 - Status: Granted - Kind: B2 - Application number: 18/108,584 - Filed: 2023-02-11 - Published: 2024-01-18 - Granted: 2025-02-04 - Earliest claimed: 2016-10-05 - Assignees: Ontocord, LLC - Applicants: Ontocord, LLC - Inventors: Hiep Huu Nguyen - Web page: https://patentgrep.com/patent/US12216664B2 ## CPC - G06F 16/24575: {using context} - G06F 40/216: using statistical methods - G06F 16/635: Filtering based on additional data, e.g. user or group profiles - G06F 40/211: Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars - G06F 40/247: Thesauruses; Synonyms - G06F 40/268: Morphological analysis - G06F 40/284: Lexical analysis, e.g. tokenisation or collocates - G06F 40/295: Named entity recognition - G06F 40/30: Semantic analysis - G06N 3/042: Knowledge-based neural networks; Logical representations of neural networks - G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] - G06N 3/0499: Feedforward networks - G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning - G06N 3/09: Supervised learning - G06N 3/098: Distributed learning, e.g. federated learning - G06N 5/025: Extracting rules from data - G06N 5/041: {Abduction} - G06N 5/046: Forward inferencing; Production systems - G06N 20/00: Machine learning - G06N 3/02: Neural networks - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06N 3/084: Backpropagation, e.g. using gradient descent - G06N 3/088: Non-supervised learning, e.g. competitive learning - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Briefly stated, the invention is directed to retrieving a semantically matched knowledge structure. A question and answer pair is received, wherein the answer is received from a query of a search engine. A question is constraint-matched with the answer based on maximizing a plurality of constraints, wherein at least one of the plurality of the constraints is a similarity score between question and answer, wherein the constraint matching generates a matched sequence. For one or more answer sequences, a subsequence is found that are not parsed as answer slots. Query results are obtained from another search engine based on a combination of the answer or question, and the non-answer subsequence. And a KB based is refined on the query results and the constraint matching and based on a neural network training, for a further subsequent semantic matching, wherein the KB includes a dense semantic vector indication of concepts. ## Claims ### Claim 1 A method for generalizing a parser for discourse assistance, comprising: receiving a text pair that includes at least a first text sequence and a second text sequence; receiving a parser that is to be generalized; receiving at least one constraint that is associated with a manner of matching text; constraint matching the first text sequence with the second text sequence using a plurality of constraints, wherein at least one of the plurality of the constraints is based on a similarity between a pair of text sequences; extracting a frame-slot of a matched sequence based on the constraint matching of the first text sequence with the second text sequence; generalizing the parser based on the generated frame-slot of the matched sequence; and returning the parser for use in an automated conversation generation system for the discourse assistance. ### Claim 2 (depends on 1) The method of claim 1, wherein the parser is generalized to perform shallow semantic parsing. ### Claim 3 (depends on 1) The method of claim 1, wherein the parser is configured to provide at least one automated conversation generation response based on extracting semantic information from a user input. ### Claim 4 (depends on 1) The method of claim 1, wherein the parser is configured to parse a user input based on the frame-slot and at least one of a location, a time, or a social network. ### Claim 5 (depends on 1) The method of claim 1, wherein the first text sequence comprises at least one of an electronic form of a written language, a spoken language, or a visually recognized sign language. ### Claim 6 (depends on 1) The method of claim 1, wherein the automated conversation generation system comprises a chat-bot system. ### Claim 7 (depends on 1) The method of claim 1, wherein the parser is usable for at least one of parsing a third text sequence for at least one of a name of an entity, a name of a product, or a name of a service. ### Claim 8 (depends on 1) The method of claim 1, wherein the generated frame-slot comprises matching word senses between the first text sequence and the second text sequence. ### Claim 9 A device for generalizing a semantically matched knowledge structure, comprising: at least one memory and at least one processor, wherein the at least one memory and the at least one processor are respectively configured to store and execute instructions for causing the device to perform operations, the operations comprising: receiving a first text sequence; receiving at least one constraint that is associated with a manner of matching text; constraint matching the first text sequence with a second text sequence based on the at least one constraint; generating a matched text sequence based on the constraint matching of the first text sequence with the second text sequence; generalizing the semantically matched knowledge structure based on the matched text sequence; and returning the generalized semantically matched knowledge structure for use in an automated conversation generation system. ### Claim 10 (depends on 9) The device of claim 9, wherein the automated conversation generation system includes a parser that is generalized for performing shallow semantic parsing with the generalized semantically matched knowledge structure. ### Claim 11 (depends on 10) The device of claim 10, wherein the parser is configured to provide at least one automated conversation generation response based on extracting semantic information from a user input. ### Claim 12 (depends on 10) The device of claim 10, wherein the parser is configured to parse a user input based on a frame-slot and at least one of a location, a time, or a social network. ### Claim 13 (depends on 9) The device of claim 9, wherein the first text sequence comprises at least one of an electronic form of a written language, a spoken language, or a visually recognized sign language. ### Claim 14 (depends on 9) The device of claim 9, wherein the automated conversation generation system comprises a chat-bot system. ### Claim 15 (depends on 10) The device of claim 10, wherein the parser is usable for at least one of parsing a third text sequence for at least one of a name of an entity, a name of a product, or a name of a service. ### Claim 16 (depends on 9) The device of claim 9, wherein the at least one processor includes a graphics processing unit. ### Claim 17 A method for generalizing a semantically matched knowledge structure, comprising: receiving a first text sequence; receiving at least one constraint that is associated with a manner of matching text; constraint matching the first text sequence with a second text sequence based on the at least one constraint; generating a matched text sequence based on the constraint matching of the first text sequence with the second text sequence; generalizing the semantically matched knowledge structure based on the matched text sequence; and returning the generalized semantically matched knowledge structure for use in an automated conversation generation system. ### Claim 18 (depends on 17) The method of claim 17, wherein the automated conversation generation system includes a parser that is generalized for performing shallow semantic parsing with the generalized semantically matched knowledge structure. ### Claim 19 (depends on 18) The method of claim 18, wherein the parser is configured to provide at least one automated conversation generation response based on extracting semantic information from a user input. ### Claim 20 (depends on 18) The method of claim 18, wherein the parser is configured to parse a user input based on a frame-slot and at least one of a location, a time, or a social network.