基于用户类别兴趣偏好的个性化排序方法  

Personalized Ranking Method Based on User Preferences of Classes Interests

作　　者：吴谈 周栋[1] 包恒泽 Wu Tan;Zhou Dong;Bao Hengze(School of Computer Science and Engineering,Hunan University of Science and Technology,Xiangtan 411201,China)

机构地区：[1]湖南科技大学计算机科学与工程学院,湖南湘潭411201

出　　处：《湖南科技大学学报（自然科学版）》2020年第1期104-112,共9页Journal of Hunan University of Science And Technology：Natural Science Edition

基　　金：国家自然科学基金资助项目(61876062);湖南省自然科学基金资助项目(2017JJ2101);湖南省教育厅科研项目资助(16K030)。

摘　　要：信息检索中,个性化排序在传统的基于内容匹配的排序算法基础上,结合用户兴趣特征,返回更符合用户需求的检索结果.由于用户数据存在稀疏性和兴趣爱好不均衡等问题,用户兴趣偏好模型构建通常不是很精确,检索效果也不佳.本文在前人研究的基础上,提出了一种基于用户类别偏好的个性化排序方法.该方法首先借助词向量技术计算查询词和文档标签集之间的语义相似程度,其次,考虑到用户对不同兴趣的偏好程度不一,通过构建用户兴趣偏好模型,计算出用户对不同兴趣类别的偏好程度,对待查询文档进行个性化处理,以达到个性化排序的目的.在真实数据集上的实验表明,与传统方法相比,本文提出的方法可以有效地改善用户的个性化检索效果.In information retrieval,the method of personalized sequencing based on the traditional content matching sorting algorithm combined with the interest of users in order to achieve some better search results that were more in line with users’needs.Due to the sparseness of user’s data and the imbalance of interests,the construction of user’s interest preference model was usually not very accurate,and the retrieval effect was little and poor.Based on those previous studies,a personalized ranking method,based on user category preference,was proposed.Firstly,the word embedding technology was used to calculate the semantic similarity between the query words and the document tag set.Secondly,considering the user’s preference for different interests,the user interest model was constructed to calculate the user’s interest categories.The degree of preference,the query documents were personalized to achieve the purpose of personalized sorting.Experiments on real data sets show that compared with traditional methods,the proposed method effectively improve the user’s personalized retrieval results.

关 键 词：个性化排序 社会化标注 词向量 兴趣偏好模型 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]