摘要： 基于TF-IQF模型的建议方法不考虑用户查询行为的上下文，在满足用户个性化需求方面存在缺陷。针对这一情况，在该方法的基础上进行优化改进，根据不同用户的查询上下文来分析用户的查询偏好，重新排序系统推荐的查询。实验结果表明，改进方法能够给出个性化的查询建议，提高用户查询的满意度。

关键词: 查询建议, 个性化, TF-IQF模型

Abstract: The query recommendation methods by mining search engines query logs can improve user satisfaction, such as the typical method based on TF-IQF model, however it does not take into account the context of user query behavior and can not meet the needs of individual users. In this paper, an improved query recommendation method based on TF-IQF model is proposed by analyzing user's query preference in the context of user queries and reordering the system recommended query. Experimental results show the improved query recommendation method gives personalized query suggestions and improves user query satisfaction further.

Key words: query recommendation, personalized, TF-IQF model

中图分类号: 

• N945