JISE


  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]


Journal of Information Science and Engineering, Vol. 35 No. 1, pp. 61-86


On Identifying Cited Texts for Citances and Classifying Their Discourse Facets by Classification Techniques


JEN-YUAN YEH1,+, TIEN-YU HSU1, CHENG-JUNG TSAI2,
PEI-CHENG CHENG3 AND JUNG-YI LIN4
1Department of Operation, Visitor Service, Collection and Information Management
National Museum of Natural Science
Taichung, 404 Taiwan
E-mail: {jenyuan; dan}@mail.nmns.edu.tw

2Department of Mathematics
National Changhua University of Education
Changhua, 500 Taiwan
E-mail: cjtsai@cc.ncue.edu.tw

3Department of Information Management
Chien Hsin University of Science and Technology
Taoyuan, 320 Taiwan
E-mail: pccheng@uch.edu.tw

4AI Lab, Semiconductor Business Group
Hon-Hai Technology Group (Foxconn)
Taipei, 114 Taiwan
E-mail: jungyilin@gmail.com


Creating the faceted citation summary of a research paper involves identifying cited texts for citation sentences (i.e., citances), classifying their discourse facets, and generating a structured summary from the cited texts. This paper proposes a supervised method for the first two tasks by classification techniques. The first task uses binary classification to distinguish relevant pairs of citances and reference sentences from irrelevant pairs. The second task applies multi-class classification to assign one of the predefined discourse facets to relevant pairs of the first task. The proposed method is evaluated using the CLSciSumm 2016 datasets and found to be competitive in producing superior results compared to state-of-the-art methods.


Keywords: citation analysis, citation linkage identification, discourse facet classification, binary/multi-class classification, scientific paper summarization

  Retrieve PDF document (JISE_201901_04.pdf)