基于句子聚类的中文文本自动摘要算法的研究

摘要

Automatic text summarization has a wide application in many fields,such as search engine and news content recommendation.The classic text summarization algorithm is to extract the keywords in the text,which ignores the relevance between the sentences in the text,and the extracted keywords are usually lack of semantic and grammatical relevance.The text is divided by sentences,sentences for clustering,divides the text into a number of relatively fixed semantic units,each unit of semantic core words,finally the core word combination of each sentence semantic construction of text summarization,test results show that the improved automatic text summarization algorithm can more effectively recall the theme of the text.%文本自动摘要在搜索引擎和新闻内容推荐等多个领域都有着非常广阔的应用.经典的文本摘要算法是提取文本中关键词进行重组,这种方式忽略了文本中句子之间的关联性,而且提取出的关键词通常缺乏语义和语法上关联性.提出了将文本以句子进行划分,针对句子进行聚类,将文本划分为一定数量语义相对固定的单元,对每个语义单元进行核心词发现,最后组合各个语义句子的核心词构建文本摘要,试验结果表明,改进的文本自动摘要算法能够更有效地召回文本主题.