Integration of domain information into state transitions of a finite state transducer for natural language processing
First Claim
1. A computer implemented method for integrating domain information and semantic parsing to generate meaning representations from natural language input, the method being implemented on a computer system having one or more physical processors programmed with computer program instructions to perform the method, the method comprising:
- receiving, by the computer system, a natural language input of a user comprising a natural language utterance in which at least a first input token has been uttered;
providing, by the computer system, the natural language input to a speech-to-text recognizer;
obtaining, by the computer system, one or more words of the natural language input as an output of the speech-to-text recognition recognizer, wherein the one or more words includes the first input token;
obtaining, by the computer system, a semantic grammar that includes word combinations for intent processing and integrates a plurality of domain tokens relating to an information domain, wherein the semantic grammar integrates the plurality of domain tokens structured into a domain information Finite State Transducer (FST) parser that includes at least a first FST path comprising a first set of domain tokens and a second FST path comprising a second set of domain tokens;
comparing, by the computer system, the plurality of domain tokens that match the first input token;
generating, by the computer system, a first score for the first FST path and a second score for the second FST path based on the comparison;
selecting, by the computer system, the first FST path based on the first score and the second score;
determining, by the computer system, a semantic structure of the one or more words based on the selected first FST path; and
generating, by the computer system, a representation of an intention of the user based on the semantic structure, wherein the representation is used to execute a natural language based search request or a natural language based command.
7 Assignments
0 Petitions
Accused Products
Abstract
The invention relates to a system and method for integrating domain information into state transitions of a Finite State Transducer (“FST”) for natural language processing. A system may integrate semantic parsing and information retrieval from an information domain to generate an FST parser that represents the information domain. The FST parser may include a plurality of FST paths, at least one of which may be used to generate a meaning representation from a natural language input. As such, the system may perform domain-based semantic parsing of a natural language input, generating more robust meaning representations using domain information. The system may be applied to a wide range of natural language applications that use natural language input from a user such as, for example, natural language interfaces to computing systems, communication with robots in natural language, personalized digital assistants, question-answer query systems, and/or other natural language processing applications.
902 Citations
25 Claims
-
1. A computer implemented method for integrating domain information and semantic parsing to generate meaning representations from natural language input, the method being implemented on a computer system having one or more physical processors programmed with computer program instructions to perform the method, the method comprising:
-
receiving, by the computer system, a natural language input of a user comprising a natural language utterance in which at least a first input token has been uttered; providing, by the computer system, the natural language input to a speech-to-text recognizer; obtaining, by the computer system, one or more words of the natural language input as an output of the speech-to-text recognition recognizer, wherein the one or more words includes the first input token; obtaining, by the computer system, a semantic grammar that includes word combinations for intent processing and integrates a plurality of domain tokens relating to an information domain, wherein the semantic grammar integrates the plurality of domain tokens structured into a domain information Finite State Transducer (FST) parser that includes at least a first FST path comprising a first set of domain tokens and a second FST path comprising a second set of domain tokens; comparing, by the computer system, the plurality of domain tokens that match the first input token; generating, by the computer system, a first score for the first FST path and a second score for the second FST path based on the comparison; selecting, by the computer system, the first FST path based on the first score and the second score; determining, by the computer system, a semantic structure of the one or more words based on the selected first FST path; and generating, by the computer system, a representation of an intention of the user based on the semantic structure, wherein the representation is used to execute a natural language based search request or a natural language based command. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 25)
-
-
13. A system for integrating domain information and semantic parsing to generate meaning representations from natural language input, the system comprising:
-
a computer system comprising one or more physical processors programmed with computer program instructions to; receive a natural language input of a user comprising a natural language utterance in which at least a first input token has been uttered; provide the natural language input to a speech-to-text recognizer; obtain one or more words of the natural language input as an output of the speech-to-text recognizer, wherein the one or more words includes the first input token; obtain a semantic grammar that includes word combinations for intent processing and integrates a plurality of domain tokens relating to an information domain, wherein the semantic grammar integrates the plurality of domain tokens structured into a domain information Finite State Transducer (FST) parser that includes at least a first FST path comprising a first set of domain tokens and a second FST path comprising a second set of domain tokens; compare the plurality of domain tokens with the first input token; and generate a first score for the first FST path and a second score for the second FST path based on the comparison; select the first FST path based on the first score and the second score; determine a semantic structure of the one or more words based on the selected first FST path; generate a representation of an intention of the user based on the semantic structure, wherein the representation is used to execute a natural language based search request or a natural language based command. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24)
-
Specification