融合句子情感和主题相似性的中文新闻文本情感摘要

发布时间：2019-03-26 12:52

【摘要】：新闻文本情感摘要是指通过提炼、浓缩而产生表达文本全局情感意见的摘要,旨在帮助人们快速获取文本的情感倾向。现有的文本摘要方法仅考虑主题及句子特征等因素,无法获取带有情感意见的文本摘要。针对这一问题,提出了融合句子情感和主题相似性的中文新闻文本情感摘要。首先,对文本中的句子进行情感标注;然后,在LexRank算法中加入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,再综合以上步骤的结果得到最终的情感摘要。实验结果表明,在ROUGE-1、ROUGE-2和ROUGE-W三个指标上,该方法比传统的LexRank算法均有提升,证明了同时考虑情感信息和主题信息能够更加有效地生成体现文本主要观点、情感的情感摘要。
[Abstract]:The emotional summary of news text is a summary which can express the whole emotional opinion of the text by abstracting and concentrating. It aims to help people get the emotional tendency of the text quickly. The existing text summary methods only consider the topic and sentence features, and can not get the text summary with emotional opinions. In order to solve this problem, an emotional summary of Chinese news text is proposed, which combines sentence emotion and subject similarity. Firstly, the sentences in the text are labeled with emotion, and then the sentence similarity is calculated by adding emotional information into the LexRank algorithm. Finally, the similarity between the sentence and the title is calculated according to the particularity of the news title, and the final emotional summary is obtained by synthesizing the results of the above steps. The experimental results show that the proposed method is better than the traditional LexRank algorithm in terms of ROUGE-1,ROUGE-2 and ROUGE-W. It is proved that both emotional information and subject information can be taken into account more effectively to generate the main points of view of the embodied text. An emotional summary of emotions.
【作者单位】：南华大学计算机科学与技术学院;
【基金】：国家自然科学基金资助项目(61402220,61502221)
【分类号】：TP391.1

【相似文献】