基于条件随机场的中文自动文摘系统

Automatic summarization system of a Chinese text based on conditional random field model

提出了一种基于条件随机场的中文自动文摘方法.用条件随机场来建立词性标注模型.在文摘句抽取时,引入了关键词抽取技术抽取文摘句.在生成文摘时,采用了基于规则的方法去除文摘中的冗余信息,使最后生成的文摘更具有可读性.实例表明该方法能够适应于许多领域,得到了很好的应用效果.

A automatic summarizing method of Chinese texts based on conditional random field is proposed.The conditional random field is used for establishing part of speech tagging model.Key-word extraction is used in the extraction of abstract sentences.The redundant information in the generated abstract is removed based on a series of rules,which can enhance the readability of the final generated abstract.Several cases show that this abstract extraction method has good application results to many types of Chinese texts.