中文摘要：

个性化搜索引擎是指在普通搜索引擎的基础上，根据用户的背景，兴趣等调整排序算法，针对不同的用户提供不同的服务。本文对搜索引擎的排序算法和用户行为进行了深入细致的研究，通过隐性方法收集用户行为信息，统计并构建用户长期兴趣模型，短期兴趣模型，时段兴趣模型等，利用向量相似度计算获取最适合用户当前状态的兴趣模型，最后将该兴趣模型结合到搜索引擎的排序算法中，影响文档得分，实现结果集的个性化排序。实验证明，该方法简单有效，可以增加搜索引擎对用户兴趣的应变能力。

英文摘要：

Personalized search engine serves individuals with diverse answers by using user interest to resort pages offered by normal search engine. Based on study of user interest model, the classification of user interest is induced into re-ranking method, which divides user interest into Long-Term Interest, Short-Term Interest and Timed Interest. The final score of the document is influenced by the most similar Interest Model, results in the personal ranking. The results show that it is simple and efficient, and can enhance the system with the ability to meet the needs, especially when the user interest is changed.