Abstract: The paper firstly reviews existing methods of relevance feedback. Then, a new implicit feedback method for building and modifying user's profiles model is presented. Finally, the implementation and experiment results of InfoAgent based on implicit feedback is introduced. The experiment demonstrates that implicit feedback is helpful to improve retrieval precision.

摘要： 回顾了已有的相关反馈技术，在此基础上提出了构造和调整用户兴趣模型的隐式反馈算法，给出了一个基于隐式反馈的InfoAgent的设计实现和实验结果。实验表明隐式反馈技术对提高检索精度有很大的帮助。

关键词: 信息检索, 隐式反馈, 相关反馈, 用户兴趣模型, 强化学习算法

CLC Number: 

• TP391.3