Research On Identification Of Professional Fake Reviewers Based On The User's Characteristics

With the rapid progress of the internet age,we are surrounded by huge amounts of Internet information,and this information plays an increasingly important role in life.User review data become the most important reference information when buyer making decision.It can be said that the quality of the reviews can make direct impact on the decision.But fake reviews interfere with buyer' view,greatly affect buyers'experience.Therefore,if through a certain method,we can recognize and filter the fake reviews and professional fake reviewers on the e-commerce platform and professional false commentators,this will greatly improve consumer shopping experience.This field has become a hot area in academic research,which attracts the attention of domestic and foreign researchers.In this article,the differences between abnormal and normal sellers,professional false commentators and normal users are analyzed,we establish a recognition model of abnormal sellers and professional false commentators.The false identification of professional false reviewers is based on the degree of network similarity,some meaningful results are achieved.Specifically,this paper includes the following work:(1)we studied all differences between abnormal and normal sellers,professional false commentators and normal users,and defined two type characteristics of sellers and the buyers.An abnormal seller analysis with seller features are constructed,which is used to locate the set of suspicious items that most likely contain professional false commentators.A model of identifying professional false commentators is constructed by using buyer's characteristics.The effectiveness evaluation of feature recognition is conducted,based on real collected data from Taobao,and results of feature recognition are given.(2)In this paper,we construct the method of network similarity combining the social relation to analyse the user similarity calculated by the user characteristic,and identify the false social commentators by the social relation network subgroup classification method.And the recognition experiment is carried out on the real set of collected data from Taobao,the recognition effects are compared.The content of this paper is helpful to the detection and identification of false professional commentators.It can be used as a supplement to the existing research in this field,which can provide some inspiration for the management of e-commerce websites.