Natural language parser
OwnerDavid N. Weiseinventor
Life of this patent
Estimating the term…
Abstract
An improved natural language parser uses a template set to identify predefined parse conditions and to implement responsive actions, thus reducing processing time while increasing parse tree accuracy. The template set is used to identify predefined conditions of the current parse state, an input span, and the history of the parse process. Once a template has been matched, the improved parser implements associated responsive actions to reduce the parse time typically associated with parsing an input span under the identified conditions. This on-the-fly implementation of responsive actions permits the parser to focus the parse process on the portions of the input span most likely to generate the complete and accurate parse in the shortest amount of time.
Claims36
A method of parsing a natural language input span, the method comprising the steps of:
- breaking the input span into individual words;
- looking up each individual word in a lexicon to determine at least one part of speech associated with the word;
- storing each individual word and an associated part of speech as a node in a candidate list;
- assigning a goodness measure to each node in the candidate list;
- setting an anchor point to a right-most portion of the input span;
- creating a grow node that contains the anchor point;
- promoting the grow node to a node chart;
- determining whether the grow node and a first node satisfy a syntax rule;
- if the grow node and the first node do not satisfy the syntax rule, comparing a parse condition to a first template;
- and
- if the parse condition and the first template match, performing a first responsive action associated with the first template.
A computer readable medium on which is stored computer-executable instructions for performing the steps of:
- breaking an input span into a plurality of words;
- associating each word with a part of speech;
- storing each word and each associated part of speech as a leaf node;
- creating a grow node that contains a right-most portion of the input span;
- growing the grow node leftward;
- if the grow node cannot be grown leftward, comparing a parse condition to a first template;
- and
- if the parse condition and the first template match, performing a first responsive action.
A method of increasing efficiency of a natural language parser for parsing an input span, the method comprising the steps of:
- creating a grow node that contains a right-most portion of the input span;
- growing the grow node leftward;
- if the grow node cannot be grown leftward, comparing a parse condition to a template;
- and
- if the parse condition and the template match, performing a first responsive action.
Dependencies are inferred from the claim text.