Abstract

Question answering (QA) is a natural language processing application in which, given a question in plain natural language and a document collection, a computer system delivers an exact answer (i.e., without spurious text) to that question, based on a document collection. General QA systems are open domain and there are no restrictions on the question type or topic. Advanced QA systems are highly complex, in that they typically involve many underlying technologies in natural language processing. Widespread research in open-domain QA started in the late 1990s and was fostered by the introduction in 1999 of a QA track at each international text retrieval conference at the U.S. National Institute of Standards and Technology, resulting in rapid advancements in the field. The major components of a QA system are question analysis, information retrieval, and answer generation.

Date

2006

Authors

U Hermjakob

Pages

323-327

Publisher

Elsevier