一种基于中文Deep Web的属性相似度计算方法

随着中文网站数量的日益庞大，中文的Deep Web信息集成已成为网络信息领域的焦点。属性相似度计算是Deep Web信息集成中模式匹配的关键一步。该文根据中文查询接口中的属性词汇常常表现为1-3个词语构成的短语的特点，提出一种更有效的基于《知网》的属性相似度计算的方法，使属性匹配的准确率得到大幅度提高。

A Calculation Method of Attribute Similarity based on Deep Web in Chinese

With the increasing amount of the Chinese website, Deep Web information integration in Chinese has become the focus of the net information field. The Attribute similarity computation is a key step of schema matching in Deep Web information integration. In this passage ,According to the characteristics of Chinese query interface attribute vocabulary which often shows of 1-3 words to form phrases, provides a more effective attribute similarity calculating method based on the “HowNet”, in this way, the attribute matching accuracy has been greatly improved.