Alignment-based surface patterns for factoid question answering systems

Issue title: Selected papers from the IEEE Conference on Information Reuse and Integration (IRI), July 13–15, 2008

Guest editors: S.H. Rubin and S.-C. Chen

Article type: Research Article

Authors: Sung, Cheng-Lung^{a; b; *} | Lee, Cheng-Wei^b | Yen, Hsu-Chun^a | Hsu, Wen-Lian^b

Affiliations: [a] Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan | [b] Institute of Information Science, Academia Sinica, Taipei, Taiwan

Correspondence: [*] Address for correspondence: Prof. Wen-Lian Hsu, Institute of Information Science, Academia Sinica, Taipei, Taiwan. Tel.: +886 2 27883799 ext. 1804; Fax: +886 2 27824814; E-mail: clsung,aska,[email protected]; [email protected].

Abstract: We propose an alignment-based surface pattern approach, called ABSP, which integrates semantic information into syntactic patterns for question answering (QA). ABSP employs a new strategy to extract surface patterns from non-segmented passages. It uses the surface patterns to extract important terms from questions, and then constructs the terms' relations from sentences in the corpus. Finally, the relations are used to rank answer candidates. Our experiments show that ABSP is highly accurate, and it can be incorporated into other QA systems that have high coverage. It can also be used in cross-lingual QA systems. The approach is robust and portable to other domains.

DOI: 10.3233/ICA-2009-0313

Journal: Integrated Computer-Aided Engineering, vol. 16, no. 3, pp. 259-269, 2009