Seminar Nasional Aplikasi Teknologi Informasi (SNATI)
2004

Sistem Pengidentifikasi Otomatis Keterkaitan Topik antar Paragraf dalam Dokumen Ekspositori

Rila Mandala (Unknown)
Andreas Prasetia (Unknown)
Rinaldi Munir (Unknown)
Harlili Harlili (Unknown)



Article Info

Publish Date
09 Nov 2009

Abstract

This paper conducted on identifying topical coherency between paragraphs of anexpository text. Expository text is a kind of texts which is intended for publicity such asjournal, news etc. Many expository texts consist of long sequences of paragraphs with verylittle structural demarcation, while others consist of sequences of paragraph which stilldiscuss the same topic. This paper uses Similarity Coefficient to identify topical coherencybetween paragraphs. Similarity coefficient measures lexical similarity between adjacentparagraphs, with assumption that the more similar two adjacent paragraphs are, the morelikely it is that the current topic continues. Pengidentifikasi Keterkaitan Topik Antar Paragraf(PKTAP) is a software for identifying topical coherency between paragraphs of an expositorytext, which is developed based on this method. Topical coherency identification made byPKTAP is compared against judgments made by human readers. The result shows that thereaders have a high tendency to agree with PKTAP.Keywords: expository text, topical coherency, similarity coefficient

Copyrights © 2004