融合句子情感和主题相似性的中文新闻文本情感摘要

新闻文本情感旨在对带有情感信息的文本,通过提炼、浓缩而产生表达文本全局情感意见的情感,用以帮助人们快速获取文本的情感倾向和主旨内容。文章主要针对网络上的海量新闻文本,研究单文档的情感。首先,该方法对文本中的句子进行情感标注。其次,通过LexRank算法与情感信息计算句子相似度。最后,根据新闻标题的特殊性计算句子与标题的相似性,综合以上考虑得到情感。实验结果表明,相比于原有的LexRank算法,该文采用的方法能够更加有效的生成带有情感色彩并且最能表达文章主旨内容的情感。

Integrating Sentences Opinion And Topic Similarities for Chinese News Text Summary Of Opinion

News opinion summary arms to refine the text with emotional information, and produce opinions express emotion .It helps people to quickly access the theme content and tendency of opinions. The paper mainly aims at the massive news text on the network, research single document summary of opinion. First, the method of this paper marked sentence with opinion information. Secondly, by the LexRank algorithm and the opinion information computing sentence similarity. Finally, according to the special characteristics of the news title, the similarity between the sentence and the title is calculated and by the above consideration to get opinion summary. The results of experimention show ,compared with the original LexRank algorithm ,adopts the method can more effectively generate opinion summary with emotional colors and the best expression of substance content.