This paper examines the role of natural language (NL) processing ininformation retrieval in the context of large operational informationretrieval systems and services. State-of-the-art information retrieval systemscombine the functional capabilities of the conventional invertedfile Boolean logic term adjacency approach commonly employed bycommercial search services, with statistical-combinatorial techniques pioneeredin experimental information retrieval (IR) research, and formalnatural language processing methods and tools borrowed from artificialintelligence (AI). The emergence and ever increasing importance of end usersearching provides challenging opportunities for the integration ofsophisticated natural language analysis and processing techniques in userfriendly interfaces.
展开▼