# US 10,769,376 B2 · Domain-specific lexical analysis > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,769,376 B2 - Document ID: US10769376B2 - Status: Granted - Kind: B2 - Application number: 15/803,158 - Filed: 2017-11-03 - Published: 2019-02-21 - Granted: 2020-09-08 - Earliest claimed: 2017-08-17 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Branimir K. Boguraev; Esme Manandise; Benjamin P. Segal - Web page: https://patentgrep.com/patent/US10769376B2 ## CPC - G06F 40/284: Lexical analysis, e.g. tokenisation or collocates - G06F 40/205: Parsing - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract A method includes performing, at a device, an analysis of a domain-specific corpus to identify a base term and a modifier term. The modifier term modifies the base term in at least a portion of the domain-specific corpus. The method also includes accessing, by the device, a first entry in lexicon data. The first entry includes core data corresponding to domain-independent lexical information for the base term. The method further includes adding, based on the analysis, non-core data to the first entry. The non-core data corresponds to domain-specific lexical information for the base term. The non-core data identifies the modifier term as a domain-specific modifier of the base term. ## Claims ### Claim 1 A method comprising: initiating, at an electronic device, an analysis of a domain-specific corpus associated with a particular domain to: identify the particular domain; and identify a base term and a modifier term, the base term and the modifier term identified independently from the particular domain, wherein the modifier term: (i) modifies the base term in at least a portion of the domain-specific corpus, and (ii) is designated as a modifier term based, at least in part, upon a co-occurrence statistics that indicate that an identified term appears to modify the base term a minimum number of times; accessing, by the electronic device, a first entry in lexicon data, wherein the lexicon data is accessible by the electronic device prior to the initiating and is configured for use at the electronic device in a language processing operation, the first entry including core data corresponding to domain-independent lexical information for the base term; adding, based on the analysis, non-core data to the first entry, the non-core data corresponding to domain-specific lexical information for the base term, wherein the non-core data identifies the modifier term and the particular domain; determining that a first portion of the non-core data is in a specific position relative to the identified base term; responsive to the determination that the first portion of the non-core data is in the specific position relative to the identified base term, generating a first collocation rule; and responsive to the generation of the first collocation rule, parsing, with the first collocation rule, the first entry in lexicon data to determine a first future placement for the first portion of the non-core data. ### Claim 2 (depends on 1) The method of claim 1, further comprising determining that the modifier term modifies the base term based on at least one of co-occurrence statistics or user input. ### Claim 3 (depends on 1) The method of claim 1, wherein: the modifier term includes at least one of an adjectival modifier term, a preposition modifier term, or a nominal modifier term, and the language processing operation includes processing, by a parser, a language sample based on the lexicon data. ### Claim 4 (depends on 1) The method of claim 1, further comprising: identifying, at the electronic device, the modifier term used with the base term as a preferred domain-specific modifier in the domain-specific corpus; and updating, at the electronic device, the non-core data to identify the modifier term as a preferred domain-specific modifier of the base term. ### Claim 5 (depends on 1) The method of claim 1, further comprising generating, based on the domain-specific corpus, domain-specific parsing rules for a domain-specific lexically-driven pre-parser. ### Claim 6 (depends on 1) The method of claim 1 wherein the first collocation rule includes information indicative of a set of future placements for the non-core data relative to the identified base term. ### Claim 7 (depends on 5) The method of claim 5, further comprising: performing, at the domain-specific lexically-driven pre-parser, a domain-specific analysis of an input text based on the non-core data; and providing a partially parsed and bracketed version of the input text from the domain-specific lexically-driven pre-parser to a domain-independent rule-based parser. ### Claim 8 (depends on 5) The method of claim 5, wherein software corresponding to the domain-specific lexically-driven pre-parser is provided as a service in a cloud computing environment. ### Claim 9 (depends on 1) The method of claim 1, further comprising, based on the analysis of the domain-specific corpus, updating the non-core data to identify one or more second modifier terms as one or more additional domain-specific modifiers of the base term. ### Claim 10 (depends on 1) The method of claim 1, wherein: prior to the initiating, the lexicon data is generated by the electronic device, received by the electronic device from another device, provided by a user to the electronic device, or a combination thereof, and the non-core data includes one or more additional modifier terms for the base term for a second domain that is distinct from a first domain associated with the domain-specific corpus. ### Claim 11 A method comprising: analyzing, at an electronic device that includes a parser, a domain-specific corpus associated with a particular domain to: identify the particular domain; and identify a base term and a modifier term, the base term and the modifier term identified independently from the particular domain, wherein the modifier term: (i) modifies the base term in at least a portion of the domain-specific corpus, and (ii) is designated as a modifier term based, at least in part, upon a co-occurrence statistics that indicate that an identified term appears to modify the base term a minimum number of times; accessing a first entry in lexicon data, the lexicon data accessible by the electronic device prior to the analyzing, and the first entry including core data corresponding to domain-independent lexical information for the base term; adding, based on the analyzing, non-core data to the first entry, the non-core data corresponding to domain-specific lexical information for the base term, and the non-core data identifying the modifier term and the particular domain; performing a language processing operation at the parser, based, at least in part, upon the lexicon data; and responsive to the performance of the language processing operation at the parser, generating a first collocation rule. ### Claim 12 (depends on 11) The method of claim 11, further comprising determining that the modifier term modifies the base term based on at least one of co-occurrence statistics or user input. ### Claim 13 (depends on 11) The method of claim 11, wherein the modifier term includes at least one of an adjectival modifier term, a preposition modifier term, or a nominal modifier term, and wherein the language processing operation includes processing a language sample. ### Claim 14 (depends on 11) The method of claim 11, further comprising: identifying, at the electronic device, the modifier term used with the base term as a preferred domain-specific modifier in the domain-specific corpus; and updating, at the electronic device, the non-core data to identify the modifier term as a preferred domain-specific modifier of the base term. ### Claim 15 (depends on 11) The method of claim 11, further comprising generating, based on the domain-specific corpus, domain-specific parsing rules for a domain-specific lexically-driven pre-parser. ### Claim 16 (depends on 11) The method of claim 11 wherein the first collocation rule includes information indicative of a set of future placements for the non-core data relative to the identified base term. ### Claim 17 (depends on 15) The method of claim 15, wherein the parser includes a domain-independent rule-based parser, and further comprising: performing, at the domain-specific lexically-driven pre-parser, a domain-specific analysis of an input text based on the non-core data; and providing a partially parsed and bracketed version of the input text from the domain-specific lexically-driven pre-parser to the domain-independent rule-based parser. ### Claim 18 (depends on 15) The method of claim 15, wherein software corresponding to the domain-specific lexically-driven pre-parser is provided as a service in a cloud computing environment. ### Claim 19 (depends on 11) The method of claim 11, further comprising, based on the analyzing, updating the non-core data to identify one or more second modifier terms as one or more additional domain-specific modifiers of the base term. ### Claim 20 (depends on 11) The method of claim 11, wherein: the lexicon data is previously generated by the electronic device, received by the electronic device from another electronic device, provided by a user to the electronic device, or a combination thereof; and the non-core data includes one or more additional modifier terms for the base term for a second domain that is distinct from a first domain associated with the domain-specific corpus.