NoteSum: An integrated note summarization system by using text mining algorithms

• Published in: Information sciences Vol. 513; pp. 536 - 552
• Main Authors: Wang, Hei-Chia, Chen, Wei-Fan, Lin, Chen-Yu
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.03.2020
• Subjects: E-Learning; Note summarization; Text mining; Topic identification; Text mining; Topic identification; Note summarization; E-Learning
• ISSN: 0020-0255; 1872-6291
• DOI: 10.1016/j.ins.2019.11.011

This study implemented an integrated system of Note Summarization (NoteSum) that merged with multi-users’ notes and searched for relevant information on the Internet and, slides, and textbooks to create a summary for students to learn effectively. The integrated system's framework consists of four different modules: Topic Identification Module, Supporting Material Finding Module, Content Mapping Module and Learning Material Integrating Module. Five experiments were conducted; these resulted in the following findings. First, translating notes with the assistance of topic terms could enhance translation quality. Second, when mapping contents, NoteSum performed better in a discussion-based course rather than in a technical course. Third, the Jensen-Shannon (JS) Divergence was used to assess the generated summary that performed better for the discussion-based course. Fourth, the three attributes—presence of topic terms, number of non-topic words, and ratio of the words with important parts of speech—had different effects on different subjects. Finally, we compared NoteSum with other existing summarization systems. The results indicated that the NoteSum-generated summary was closer to students’ original notes and thus resulted in better performance in readability, informativeness, and completeness. All the results confirm that our proposed NoteSum is an effective note summarization system for student learning.