摘要
技术目前已经成为计算机语言学领域的一个研究热点。本文讨论了自动摘要的定义和分类。针对自动文摘中主题句的冗余现象,提出了一种新型的自动摘要冗余处理的方法。该方法将初始文摘中的句子表示成句链.根据任意文摘句中所有特征词的激活水平、初始化水平、影响因子以及语句相干性公式,计算其与其它初始文摘中句子的相干性.去除相干性比较大的冗余句子,从而得到最终的自动摘要。
Automatic Text Summarization technology has become a hot topic in the field of computational linguistics. This article discusses the definition and classification of automatic summary. Againsting the redundancy of the topic sentences ir automatic summary, it puts forward a new method of automatic summarization, which automatically processes prolixity. This method represents sentences in initial abstract into sentence chains. Calculate its initial coherence with other sentences in initia abstract according to activation levels and initialization levels of all the feature words in every sentence in initial abstract, influence factor and statement coherence formula. Remove the sentences which have the relatively large coherence, thus get the fina automatic summarization.
出处
《中国新通信》
2014年第14期92-93,共2页
China New Telecommunications
基金
国家高技术研究发展计划(项目编码:2012AA101008)
关键词
自动摘要
冗余处理
语句相干性
automatic text summarization prolixity processing Statement coherence