Natural language information retrieval system and method

作者: Elizabeth D. Liddy , Mary E. McKenna , Woojin Paik , Ming Li

DOI:

关键词: Query expansionRanking (information retrieval)Natural language processingWeb search queryWeb query classificationInformation retrievalQuery languageRDF query languageComputer scienceQuery optimizationArtificial intelligenceQuery by Example

摘要: Techniques for generating sophisticated representations of the contents both queries and documents in a retrieval system by using natural language processing (NLP) techniques to represent, index, retrieve texts at multiple levels (e.g., morphological, lexical, syntactic, semantic, discourse, pragmatic levels) which humans construe meaning writing. The user enters query processes generate an alternative representation, includes conceptual-level abstraction based on complex nominals (CNs), proper nouns (PNs), single terms, text structure, logical make-up query, including mandatory terms. After displays information user, indicating system's interpretation representation content query. is then given opportunity provide input, response modifies Once has provided desired possibly modified matched relevant document database, measures relevance generated documents. A set presented who select some or all documents, typically basis such being particular relevance. initiates generation selected document(s).

参考文章(20)
Elizabeth D Liddy, Woojin Paik, Edmund S Yu, Mary McKenna, None, Document retrieval using linguistic knowledge multimedia information retrieval. pp. 106- 114 ,(1994)
Elizabeth D Liddy, Sung H Myaeng, None, DR-LINK: A System Update for TREC-2 text retrieval conference. pp. 85- 100 ,(1994)
E. Liddy, Sung Hyon Myaeng, DR-LINK's linguistic conceptual approach to document detection text retrieval conference. pp. 113- 129 ,(1993)
Elizabeth D. Liddy, Mary McKenna, Woojin Paik, Development and Implementation of a Discourse Model for Newspaper Texts ,(1995)