GARCH (1, 1) Outlier Detection Technique for Review Spam Detection

Siddu P. Algur ,  Jyoti G. Biradar , Prashant Bhat

Abstract With the recent development of web 2.0, the volume of on-line sales has been increasing in a remarkable pace and has generated plentiful of user-created content. Among various types of user-generated data on the web, reviews about stores, products, businesses, or services written by users are becoming more and more important due to the word-of-mouth effect and their impact on influencing customers purchase decisions. These reviews are important source of information for the potential customers before deciding to purchase a product. As a result, websites containing customer reviews are becoming targets of opinion spam. Spam reviews corrupt the online review system and confuse the consumers. Hence, a novel and effective technique GARCH (1,1) model is used in this work to find review spamicity in the multidimensional time series reviews of the stores, extracted from review website resellerratings.com. The experimental results demonstrates that the method proposed is effective in detection of review spamicity. Keywords:- Review spam, outliers, time series, multidimension, GARCH.

Title:	GARCH (1, 1) Outlier Detection Technique for Review Spam Detection
Author Name:	Siddu P. Algur , Jyoti G. Biradar , Prashant Bhat
Abstract:	Abstract With the recent development of web 2.0, the volume of on-line sales has been increasing in a remarkable pace and has generated plentiful of user-created content. Among various types of user-generated data on the web, reviews about stores, products, businesses, or services written by users are becoming more and more important due to the word-of-mouth effect and their impact on influencing customers purchase decisions. These reviews are important source of information for the potential customers before deciding to purchase a product. As a result, websites containing customer reviews are becoming targets of opinion spam. Spam reviews corrupt the online review system and confuse the consumers. Hence, a novel and effective technique GARCH (1,1) model is used in this work to find review spamicity in the multidimensional time series reviews of the stores, extracted from review website resellerratings.com. The experimental results demonstrates that the method proposed is effective in detection of review spamicity. Keywords:- Review spam, outliers, time series, multidimension, GARCH.
Cite this article:	Siddu P. Algur , Jyoti G. Biradar , Prashant Bhat , " GARCH (1, 1) Outlier Detection Technique for Review Spam Detection " , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS), Volume 5, Issue 6, November - December 2016 , pp. 006-015 , ISSN 2278-6856.
Full Text [PDF] Home