Question Answering Supported By Multiple Levels Of Information Extraction

This chapter discusses the importance of information extraction (IE) in question answering (QA) systems. Most QA systems are focused on sentence-level answer generation. Such systems are based on information retrieval (IR) techniques such as passage retrieval in conjunction with shallow IE techniques such as named entity tagging. Sentence-level answers may be sufficient in many applications focused on reducing information overload: instead of a list of URLs provided by search engines which need further perusal by the user, sentences containing potential answers are extracted and presented to the user. However, if the goal is precise answers consisting of a phrase, or even a single word or number, new techniques for QA must be developed. Specifically, there is a need to use more advanced IE and natural language processing (NLP) in the answer generation process. This chapter presents a system whereby multiple levels of IE are utilized in a QA system that attempts to generate the most precise answer possible, backing off to coarser levels where necessary. In particular, generic grammatical relationships are exploited in the absence of information about specific relationships between entities. An IE engine, InfoXtract, is first described to illustrate the types of IE and NLP output that is needed. Results are presented for QA based on named entity tagging, relationships between entities, and semantic parsing, including more recent work that handles caseless input. This work has implications for broader coverage QA systems since domain-independent IE results can be exploited and noisy input can be “normalized” by restoration techniques.