Penjanaan Ringkasan Isi Utama Berita Bahasa Melayu berdasarkan Ciri Kata (Generation of News Headline for Malay Language based on Term Features)

Abstract

© 2018, Universiti Kebangsaan Malaysia Press. All rights reserved. Headline generation is an information extraction process to generate a single sentence that represents the content of a text. In Malay language context, research in this area is limited to machine translation approaches. This study is divided into three phases: analysis of news discourse, development of headline generation technique and evaluation of the quality of generated headlines. The study aims to develop headline using statistical and linguistic methods. The statistic method used to identify significant words and sentences based in term weighting approach. The linguistic method is used to increase its preciseness. 140 news and their corresponding headlines model were constructed. Analysis of the news collection shows that the main idea of written text can be identified based on four characteristics: word location in sentences, sentence location in texts, acronym word types and words that represent the person name. Significant words with main idea of written text are determined based on the words weighted values. The values are determined by combining the frequency of words and word location in sentences. The content of the first two sentences are suitable candidates for recognising important sentences in text. Results showed that mean percentage for important sentence recognition 82.9%, mean quality of generated headlines are 0.3194 (precision), 0.5656 (recall), 0.4012 (F-measure), 0.5656 (ROUGE–N), 0.3392 (ROUGE–L), 0.1186 (ROUGE–W) and 0.1232 (ROUGE–S). In conclusion, the consideration of language factors in headline generation technique is capable of producing quality headlines with higher degree of fidelity as compared to the compared benchmarks.