KeyphraseDS: Automatic generation of survey by exploiting keyphrase information

详细信息    查看全文

• 作者：Shansong Yang^aAuthor Vitae ; Weiming Lu^a ; ^{luwm@zju.edu.cn}Author Vitae ; Dezhi Yang^aAuthor Vitae ; Xi Li^aAuthor Vitae ; Chao Wu^bAuthor Vitae ; Baogang Wei^aAuthor Vitae
• 关键词：Document summarization ; Survey generation ; Keyphrase extraction ; Semantic similarity ; ILP optimization
• 刊名：Neurocomputing
• 出版年：2017
• 出版时间：8 February 2017
• 年：2017
• 卷：224
• 期：Complete
• 页码：58-70
• 全文大小：4156 K
• 卷排序：224

文摘

KeyphraseDS can organize scientific articles into multi-aspect and informative scientific survey by exploiting keyphrases. KeyphraseDS consists of three steps: keyphrase graph construction, semantic structure generation and sentence selection. Semantic structure is computed based on the topic-based similarity and knowledge-based similarity. Sentences are selected by taking semantic relevance, semantic diversity and keyphrase salience into consideration. An elaborate human evaluation is conducted to evaluate different methods’ performance.