Named Entity Recognition for Question Answering

Proceedings of the 2006 Australasian Language Technology Workshop (ALTW2006), pages 51–58. Named Entity Recognition for Question Answering Diego Moll´a and Menno van Zaanen and Daniel Smith Centre for Language Technology Macquarie University Sydney Australia {diego, menno, dsmith}@ics.mq.edu.au Abstract Current text-based question answering (QA) systems usually contain a named en- tity recogniser (NER) as a core compo- nent. Named entity recognition has tra- ditionally been developed as a component for information extraction systems, and current techniques are focused on this end use. However, no formal assessment has been done on the characteristics of a NER within the task of question answering. In this paper we present a NER that aims at higher recall by allowing multiple entity labels to strings. The NER is embedded in a question answering system and the over- all QA system performance is compared to that of one with a traditional variation of the NER that only allows single entity la- bels. It is shown that the added noise pro- duced introduced by the additional labels is offset by the higher recall gained, there- fore enabling the QA system to have a bet- ter chance to find the answer. 1 Introduction Many natural language processing applications re- quire finding named entities (NEs) in textual doc- uments. NEs can be, for example, person or com- pany names, dates and times, and distances. The task of identifying these in a text is called named entity recognition and is performed by a named entity recogniser (NER). Named entity recognition is a task generally associated with the area of information extrac- tion (IE). Firstly defined as a separate task in the Message Understanding Conferences (Sund- heim, 1995), it is currently being used in a varied range of applications beyond the generic task of information extraction, such as in bioinformatics, the identification of entities in molecular biology (Humphreys et al., 2000), and text classification (Armour et al., 2005). In this paper we will focus on the use of named entity recognition for question answering. For the purposes of this paper, question answering (QA) is the task of automatically finding the answer to a question phrased in English by searching through a collection of text documents. There has been an increase of research in QA since the creation of the question answering track of TREC (Voorhees, 1999), and nowadays we are starting to see the introduction of question-answering techniques in mainstream web search engines such as Google1, Yahoo!2 and MSN3. An important component of a QA system is the named entity recogniser and virtually every QA system incorporates one. The rationale of incor- porating a NER as a module in a QA system is that many fact-based answers to questions are en- tities that can be detected by a NER. Therefore, by incorporating in the QA system a NER, the task of finding some of the answers is simplified consid- erably. The positive impact of NE recognition in QA is widely acknowledged and there are studies that confirm it (Noguera et al., 2005). …

Content truncated to 500 words