Abstract: Knowledge push is a hot topic of knowledge reuse.After studying the commonly used knowledge push approaches，this paper presents a knowledge push approach based on attribute similarity.The approach analyses the knowledge set that has been viewed by user，and gets user’s interest information by calculating the attribute similarity of the knowledge set，then matches the attribute similarity of the unviewed knowledge with the interest information to judge whether the unviewed knowledge is satisfied with user’s interest.By this way，the approach implements knowledge push without manual participation.The experiment results show that this method can push knowledge efficiently.

Key words: knowledge push, attribute, similarity, knowledge management

摘要： 知识推送是知识重用的一个热点问题。在研究了现有推送技术的基础上，提出了一个基于属性相似度的知识推送方法。该方法对用户已经浏览过的知识进行分析，通过计算这些知识的属性相似度得到用户的知识兴趣信息，把用户未浏览的知识与用户知识兴趣信息进行属性相似度匹配，以判断该知识是否符合用户的知识兴趣，从而实现了无需人工干预的知识推送。实验结果表明，该方法具有较高的推送准确性。

关键词: 知识推送, 属性, 相似度, 知识管理