Research On Review Spam Detection Based On Imbalanced Data Classification Method

With the rapid development and widespread popularization of the Internet,especially the introduction of Web2.0 concept and technology,massive user generated contents have been generated on the Internet,which include online reviews about e-commerce products.However,due to the openess of the Internet and motivated by the huge benefits from online reviews,a lot of spam reviews are produced.Only manual analysis can be not capable.Therefore,it is necessary to introduce the review spam detection technique to analyze reviews.Having considered the imbalanced distribution of the normal reviews and the spam reviews,we study on the review spam detection from the perspective of imbalanced data classification.Firstly,the backgrounds and the research significances of the review spam detection are analyzed.Secondly,the basic theories about review spam detection and imbalanced data classification are comprehensively researched,which mainly includes the concept of spam review,an overview of review spam detection,the difficulty of review spam detection,the existing methods and features for review spam detection,and the overview of existing imbalanced data classification methods from data-level and algorithm-level.Then,an improved method is proposed based on SVM to deal with the class imbalance problem,besides,a novel model is constructed for the detection of spam reviews based on the imbalanced data classification methods.Finally,oriented to the application of e-commerce,a review spam detection prototype system is developed.The validity and practicability of the developed model are verified by applying it into practical applications.The experimental results verify the effectiveness of the proposed model.In this study,an improved method is proposed based on SVM to address the data imbalance problem,which have enriched and perfected the research system of review spam detection.By applying the model into e-commerce,a review spam detection prototype system is developed,which provides an effective way for enterprises to solve the problem of spam reviews.