基于二级向量描述的搜索引擎个性化服务模型  

A Search Engine Personalization Model Based on Double Vector Description

作　　者：徐静秋[1] 朱征宇[1] 谭明红[1] 任翔[1] 

机构地区：[1]重庆大学计算机学院,重庆400044

出　　处：《计算机科学》2007年第11期89-93,共5页Computer Science

基　　金：高等学校博士学科点专项科研基金资助课题(20030611016);重庆大学骨干教师计划基金项目(2003A33)

摘　　要：本文介绍了一种搜索引擎个性化服务模型。用二级向量进行文本特征提取和用户兴趣建模,关键词向量能快速定位用户的兴趣领域,而扩展词向量能准确反映用户在该领域上的兴趣偏好。当用户提交关键词时,本系统根据学习到的用户兴趣描述模型计算词间相关度,自动增加几个个性化扩展词提交给搜索引擎,实现不同用户键入相同关键词能返回不同信息的目的。实验结果充分表明本系统应用于搜索引擎个性化服务领域的有效性和实用性。A novel personalized search engine model is proposed in this article. It describes user profiles and text feature with double vector. Keyword vector helps to rapidly find a user＇s interest domain. Expansion word vector can accurately express his preference in the domain. When a user submit query keywords, our system automatically computes the term-term associations according to user profiles. More personalized expansion words are determined using a search engine, Thereby different search results are returned to different users who input the same key words. The test results show the feasibility and applicability of the presented work for personalized information service of a search engine.

关 键 词：搜索引擎 个性化 二级向量描述 词间相关度 

分 类 号：TP393.4[自动化与计算机技术—计算机应用技术]