个性化信息检索中的用户历史分析  

User Search History Analysis in Personalized Information Retrieval

作　　者：王晓春[1] 杨沐昀[1] 李生[1] 

机构地区：[1]哈尔滨工业大学机器智能与翻译实验室,哈尔滨150001

出　　处：《智能计算机与应用》2014年第5期99-102,共4页Intelligent Computer and Applications

基　　金：国家自然科学基金重点项目(61105072;61272384);国家高技术研究发展计划(863)(2011AA01A207)

摘　　要：个性化信息检索通过挖掘用户兴趣来准确理解用户信息需求,提供高质量的检索结果。然而,用户历史是否与用户信息需求有关系这一问题缺少明确回答。本文基于真实搜索日志对用户历史与用户信息需求的关系进行了分析。当前点击代表了用户信息需求。分析结果显示,用户历史中存在与用户信息需求有关的内容,证明挖掘用户历史可以提高个性化检索性能。Personalized search provides better retrieval results by mining individual interests to help understand user irdbr- mation needs accurately. However, the relationship between user search history and information needs is less touched. This paper investigates their relationship based on search engine log. The clicks in the current retrieval result represent individu- al information needs. Analysis results show that user search history contains information related to current clicks, suggesting that personalized retrieval performance can be improved by user search history.

关 键 词：个性化信息检索 用户历史 信息需求 当前点击 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]