Answerfinder: Question Answering by Combining Lexical, Syntactic and Semantic Information

AnswerFinder — Question Answering by Combining Lexical, Syntactic and Semantic Information Diego Moll´a and Mary Gardiner Centre for Language Technology Division of Information and Communication Sciences Macquarie University Sydney, Australia diego@ics.mq.edu.au and gardiner@ics.mq.edu.au Abstract We present a question answering system that combines information at the lexical, syntactic, and semantic levels, in the process to find and rank the candidate answer sentences. The can- didate exact answers are extracted from the candidate answer sentences by means of a com- bination of information-extraction techniques (named entity recognition) and patterns based on logical forms. The system participated in the question answering track of TREC 2004. 1 Introduction Question answering is an area that is becom- ing increasingly active in research and is cur- rently being deployed into practical applica- tions. Research in question answering has re- cently been fostered by large-scale programs like AQUAINT1 and evaluation frameworks like TREC2, NTCIR3, and CLEF4. Such research and the current need to cope with large vol- umes of text has led various companies to pro- duce practical question answering systems. For example, research groups from Microsoft, IBM, NTT, Oracle, and Sun have participated in the question answering track of TREC. In addition, there are several attempts to provide question- answering extensions to the current Web search engines, with demos available by MIT5, LCC6, and BrainBoost7, among others. AnswerFinder is an open-domain question an- swering system that combines information at the lexical, syntactic, and semantic levels in various stages to find the exact answer to the user question. This paper describes the An- swerFinder system as it stood at the time of the 1www.ic-arda.org/InfoExploit/aquaint/index.html 2trec.nist.gov 3research.nii.ac.jp/ntcir/index-en.html 4clef.iei.pi.cnr.it 5www.ai.mit.edu/projects/infolab/ 6www.languagecomputer.com/ 7www.brainboost.com/ TREC 2004 question answering track. Section 2 introduces TREC and the question answering track. Section 3 describes the architecture of the system. Section 4 details the function of each module within the AnswerFinder system. Section 5 gives the system performance on the TREC 2003 question set. Section 6 mentions re- lated work and Section 7 lists problems with the AnswerFinder system that should be addressed in the near future. 2 The TREC 2004 Question Answering Track The Text REtrieval Conference (TREC) started in 1992 as part of the TIPSTER text program. A fundamental goal of the conference is to pro- vide an evaluation framework for the compar- ison of the results of independent information retrieval systems. The concept of information retrieval is to be understood in a broad sense, and this conference has developed various tracks that focus on specific areas of information re- trieval, such as ad-hoc (the name given to docu- ment retrieval), routing, speech, cross-language, web, video, and very large corpora (Voorhees, 2003). The question answering track started in 1999 and ever since its creation it has been the most popular track. Every year the complexity and difficulty of the task increases. Thus, in 1999 the competing systems were asked to retrieve small snippets of text containing the answer. The questions were designed by the participants and the answer was guaranteed to be in …

Content truncated to 500 words