中文摘要：

以语义为基础实现文档关键词提取是提高自动提取准确度的有效途径。以中文文档为处理对象，通过《同义词词林》计算词语间语义距离，对词语进行密度聚类，得到主题相关类，并从主题相关类中选取中心词作为关键词。通过统计实验和打分实验，证日月基于语义的文档关键词提取方法具有较高的准确率、召回率，并且提取的关键词具有较高的主题相关度。

英文摘要：

Document keywords extraction on the basis of semantic was an effective way to improve the accuracy of automatic extraction. This paper regarded Chinese document as processing object, calculated the semantic distances between words through the synonyms dictionary. Then, through density clustering of the words, it got theme related classes. Finally, it regar- ded the headwords selected from topic related classes as keywords. Statistical experiment and scale experiment prove that the semantic-based keyword extraction method for document has higher accuracy, recall rate and the extracted keywords have high- er related degrees to the topic.