Browse Prior Art Database

A system and method for improving query results by minimizing ambiguity in search terms

IP.com Disclosure Number: IPCOM000235788D
Publication Date: 2014-Mar-25
Document File: 2 page(s) / 25K

Publishing Venue

The IP.com Prior Art Database

Abstract

A method for improving query results by minimizing ambiguity in search terms is disclosed.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 55% of the total text.

Page 01 of 2

A system and method for improving query results by minimizing ambiguity in search terms

Disclosed is a method for improving query results by minimizing ambiguity in search terms.

Terms used in queries of unstructured corpora in deep analytical question and answer systems often have multiple meanings. For example, "Buffalo" can be an animal or a city name. This can result in search hits being returned which are not related to the actual question, resulting in inaccurate or voluminous search results, and also creating performance problems.

Prior to indexing a document, natural language processing (NLP) can be run against the document, identifying the parts of speech, semantic types, etc of terms to be indexed. This information can then be applied to the indexed terms as 'field' information. If similar NLP is performed on the input question, the same information is obtained for terms to be used in unstructured queries. When constructing queries, terms would only be searched for in matching fields.

The following method is disclosed for improving searches in question answering systems:

Before a document is added to a searchable store, NLP is utilized to determine the parts of speech, semantic types, etc. of meaningful words (i.e. non stop words) in the document. When the document is added to a searchable store, the words are assigned field information which identifies the syntactic and semantic information determined.

When the deep analytical question and answer syst...