Arabic Text Summarization from Reader's Perspective

Mohammad Hjouj Btoush ； Nasim F. Diab ； Khalid N. Alzubi

《Research Journal of Applied Sciences, Engineering and Technology》 10卷6期 (2015/06) Pp. 680-687

並列摘要

The aim is to build a system that takes the source text as an input and the result will be a summery which contains sentences preserve the main theme of the source. We use a statistical approach to solve the problem of Arabic text summarization. Many researches which study Arabic text summarization problem focus on how writers of articles write, here we make a system take the advantage of how readers of an article read and make comments on article. Our approach assign scores for each segment of an article depending on the location of the sentences, the size of each sentence and term frequency on the article and comments of readers. Our system tested by 100 human evaluators, We gave each evaluator a copy of the summary produced by our system, with a question about the connectivity of ideas and sentences in ＜article, summary＞ and asked them to evaluate the summary. We asked them to answer by rejected, not-related, satisfactory, good, accepted, to test our approach which focus on readers' comments to summarize articles.

並列關鍵字

Comment relatedness ； extract ； term frequency ； text summarization