作者: Yufeng Jing , W Bruce Croft , None
DOI:
关键词: Natural language 、 Computer science 、 Probabilistic inference 、 Construct (python library) 、 Information retrieval 、 Association (object-oriented programming) 、 Thesaurus (information retrieval) 、 Variety (linguistics) 、 Artificial intelligence 、 Natural language processing
摘要: Although commonly used in both commercial and experimental information retrieval systems, thesauri have not demonstrated consistent benefits for performance, it is difficult to construct a thesaurus automatically large text databases. In this paper, an approach, called PhraseFinder, proposed collection-dependent association using full-text document collections. The can be accessed through natural language queries INQUERY, system based on the probabilistic inference network. Experiments are conducted INQUERY evaluate different types of thesauri, constructed variety