大数据网络用户浏览隐式反馈信息检索仿真  

Big Data Network User Browsing Implicit Feedback Information Retrieval Simulation

作　　者：屈娟娟[1] QU Juan-juan(Xinhua College of Sun Yat-Sen University,Guangdong Guangzhou 510520,China)

机构地区：[1]中山大学新华学院

出　　处：《计算机仿真》2019年第9期430-433,468,共5页Computer Simulation

基　　金：基于大数据的虚拟客户行为的电子商务实践教学平台建设和实践型电子商务人才培养体系的探讨(2017J009);重点课程网络(2018ZD007)

摘　　要：传统方法在对大数据网络用户进行浏览隐式反馈信息检索时,存在查全率低、准确率不高等问题。针对上述问题,提出一种基于大数据网络的用户浏览隐式反馈信息检索方法。方法通过观测用户在浏览网络页面时所选取的动作来获取隐式反馈信息,并根据这些信息建立用户兴趣更新模型。采用向量来描述用户浏览的网页文档,为各个浏览行为赋予相应的权值,通过该权值从用户的浏览行为推算出用户对某一文档的感兴趣程度,并建立基于用户浏览隐式反馈信息的用户兴趣模型,并利用用户兴趣更新模型对兴趣进行更新,提高检索精度,以此实现大数据网络用户浏览隐式反馈信息检索。实验仿真证明,与传统方法相比,所提方法较能够有效提高检索有效性、查全率、准确率。In traditional methods,recall rate and accuracy rate are low.Therefore,this paper puts forward a method to retrieve user browsing implicit feedback information based on big data network.This method obtained the implicit feedback information by observing the action selected by user in browsing web pages.On this basis,the user interest update model was established.Moreover,the method used vectors to describe the web documents browsed by user and gave corresponding weights to each browsing behavior.Based on this weight value,the user's interest in a document can be calculated from the user's browsing behavior.On this basis,our method deduced the user's interest degree on a document by the user's browsing behavior.Meanwhile,we built the user's browsing implicit feedback information and used this model to update the interest,so as to improve the retrieval accuracy.Thus,the implicit feedback information retrieval of use browse based on big data network could be achieved.Simulation results verify that,compared with the traditional methods,the proposed method can effectively improve the retrieval efficiency,recall rate and accuracy rate.

关 键 词：信息检索 大数据网络 用户浏览隐式 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]