Topic identification method for textual document

• Main Authors: Jamil, Nurul Syafidah, Ku-Mahamud, Ku Ruhana, Mohamed Din, Aniza
• Format: Article
• Language: English
• Published: JMEST 2017
• Subjects: QA76 Computer software
• Online Access: http://repo.uum.edu.my/21719/1/JMEST%204%202%202017%206643%206647.pdf; http://repo.uum.edu.my/21719/; http://www.jmest.org/wp-content/uploads/JMESTN42352037.pdf
• Institution: Universiti Utara Malaysia

Abstract— Topic identification is a crucial task for discovering knowledge from textual document. Existing methods for topic identification suffer from word counting problem as they depend on the most frequent terms in the text to produce the topic keyword.Not all frequent terms are relevant. This paper proposes a topic identification method that filters the important terms from the preprocessed text and applied term weighting scheme to solve synonym problem.A rule generation algorithm is used to determine the appropriate topics based on the weighted terms.The text document used in the experiment is the English translated Quran.The topics identified from the proposed method were compared with topics identified using Rough Set and domain experts. From the findings, the proposed topic identification method was consistently able to identify topics that are mostly close to the topics that have been given by Rough Set and the experts.The result from the comparison proved that the proposed method was able to be used to capture topics for textual documents.