Informative sentence retrieval for domain specific terminologies

  • Authors:

  • Jia-Ling Koh;Chin-Wei Cho

  • Affiliations:

  • Department of Information Science and Computer Engineering, National Taiwan Normal University, Taipei, Taiwan, R.O.C.;Department of Information Science and Computer Engineering, National Taiwan Normal University, Taipei, Taiwan, R.O.C.

  • Venue:
  • IEA/AIE'11 Proceedings of the 24th international conference on Industrial engineering and other applications of applied intelligent systems conference on Modern approaches in applied intelligence - Volume Part I
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Domain specific terminologies represent important concepts when students study a subject. If the sentences which describe important concepts related to a terminology can be accessed easily, students will understand the semantics represented in the sentences which contain the terminology in depth. In this paper, an effective sentence retrieval system is provided to search informative sentences of a domain-specific terminology from the electrical books. A term weighting model is constructed in the proposed system by using web resources, including Wikipedia and FOLDOC, to measure the degree of a word relative to the query terminology. Then the relevance score of a sentence is estimated by summing the weights of the words in the sentence, which is used to rank the candidate answer sentences. By adopting the proposed method, the obtained answer sentences are not limited to certain sentence patterns. The results of experiment show that the ranked list of answer sentences retrieved by our proposed system have higher NDCG values than the typical IR approach and pattern-matching based approach.