Natural language understanding system
First Claim
1. A method for interpreting natural language input using a computer system comprising the steps of:
- parsing said natural language input;
disambiguating meanings and structure of said natural language input;
generating a discourse representation, said discourse representation identifying said structure for said natural language input, meanings of words and relationships between components of said structure;
resolving references in said natural language input;
determining relationships that exist in said natural language input;
consulting, within each of the above steps, a naive semantic lexicon to determine the plausibility of interpretative decisions made within each of the above steps, said naive semantic lexicon having at least one entry that identifies at least one sense of a word, said entry including a reference to an ontological classification network, syntactic information, and a plurality of semantic properties.
5 Assignments
0 Petitions
Accused Products
Abstract
The present invention interprets natural language input using common sense reasoning. The invention avoids the combinatorial explosion that has occurred in other natural language understanding systems. The invention uses modules for parsing, disambiguation, formal semantics, anaphora resolution, and coherence, and a naive semantic lexicon. The naive semantic lexicon is consulted by the parsing, disambiguation, formal semantics, anaphora resolution, and coherence modules to determine whether an interpretation alternative is plausible based on the world knowledge contained in the naive semantic lexicon. The parsing module employs both a top-down and bottom-up parsing strategy. The parsing module consults the naive semantic lexicon to build a structure from natural language input that has both semantic and pragmatic plausibility. The invention uses a psychologically-motivated naive semantic ontology that provides a means for classifying concepts. The lexicon relates word senses to the ontological concepts, contains word sense-specific common sense knowledge, and connects syntactic information with the meaning of each word sense. Using the natural language understanding of the present invention, a process is used for retrieving text that includes the steps of: 1) natural language understanding of a document base, 2) natural language understanding of a text retrieval request (i.e., query), 3) a comparison of the output of steps 1 and 2.
-
Citations
30 Claims
-
1. A method for interpreting natural language input using a computer system comprising the steps of:
-
parsing said natural language input; disambiguating meanings and structure of said natural language input; generating a discourse representation, said discourse representation identifying said structure for said natural language input, meanings of words and relationships between components of said structure; resolving references in said natural language input; determining relationships that exist in said natural language input; consulting, within each of the above steps, a naive semantic lexicon to determine the plausibility of interpretative decisions made within each of the above steps, said naive semantic lexicon having at least one entry that identifies at least one sense of a word, said entry including a reference to an ontological classification network, syntactic information, and a plurality of semantic properties. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 26)
-
-
17. A method for retrieving text comprising the steps of:
-
inputting textual information to a natural language understanding module; translating, using said natural language understanding module, said textual information into a logic form which expresses objects and events in said textual information as concepts and relationships between said concepts that identify the structure expressed in said textual information; inputting a query to said natural language understanding module; translating, using said natural language understanding module, said query into a logic form which expresses objects and events in said query as concepts and relationships between said concepts that identify the structure expressed in said query; comparing said logic form of said query to said logic form of said textual information; and retrieving said textual information when said query'"'"'s logic form is similar to said textual information'"'"'s logic form.
-
-
18. A method for creating a cognitive model comprising the steps of:
-
storing textual information in storage of said computer system; inputting said textual information to a natural language understanding (NLU) module, said NLU module comprising parsing, disambiguation, formal semantics, anaphora resolution and coherence modules that consult a naive semantic lexicon having at least one entry that identifies at least one sense of a word, said entry including a reference to an ontological classification network, syntactic information and a plurality of semantic properties; generating, using said natural language understanding module, a cognitive model which expresses objects and events in said textual information as concepts and relationships between said concepts that identify the structure expressed in said textual information; and generating a concept index of said concepts in said cognitive model, said concept index is used to locate the textual information associated with concepts in said cognitive model.
-
-
19. A method for retrieving text in a computer system comprising the steps of:
-
storing in said computer system textual information; generating, using a natural language understanding module comprising parsing, disambiguation, formal semantics, anaphora resolution and coherence modules consulting a naive semantic lexicon having at least one entry that identifies at least one sense of a word, said entry including a reference to an ontological classification network, syntactic information and a plurality of semantic properties, a cognitive model of said textual information which expresses objects and events in said textual information as concepts and relationships between said concepts that identify the structure expressed in said textual information; generating, using said natural language understanding module, a cognitive model of a query which expresses objects and events in said query as concepts and relationships between said concepts that identify the structure expressed in said query; and generating a list of textual information relevant to said query based on said cognitive model of said textual information and said cognitive model of said query. - View Dependent Claims (20, 21, 22, 23, 24)
-
-
25. An article of manufacture comprising:
a computer usable medium having computer readable program code embodied therein for interpreting natural language input comprising; computer readable program code configured to cause a computer to parse said natural language input driven by the content of said natural language input and expectations about said natural language input using a naive semantic lexicon having at least one entry that identifies at least one sense of a word, said entry including a reference to an ontological classification network, syntactic information, and a plurality of semantic properties; computer readable program code configured to cause a computer to disambiguate during said parse said natural language input using said naive semantic lexicon; computer readable program code configured to cause a computer to generate a discourse representation said natural language input using said naive semantic lexicon; computer readable program code configured to cause a computer to resolve references in said natural language input using said naive semantic lexicon; and computer readable program code configured to cause a computer to determine relationships that exist in said natural language input using said naive semantic lexicon. - View Dependent Claims (27, 28, 29, 30)
Specification