PredictingTrust from User Ratings

https://doi.org/10.1016/j.procs.2012.06.036Get rights and content
Under a Creative Commons license
open access

Abstract

Trust relationships between usersinvariousonline communitiesare notoriouslyhardtomodelfor computer scientists.It canbeeasilyverifiedthattryingtoinfer trustbasedonthe socialnetworkaloneisoftenineffcient. Therefore, theavenue we explore is applying Data Mining algorithms to unearth latent relationships and patterns from background data. In this paper, we focus on a case where the background data is user ratings for online product reviews. We consider as a testinggroundalarge datasetprovidedby Epinions.comthat containsa trustnetworkaswellas userratingsforreviews on products from a wide range of categories. In order to predict trust we define and compute a critical set of features, which we show to be highly effective in providing the basis for trust predictions. Then, we show that state-of-the-art classifierscandoan impressivejobin predicting trustbasedonourextracted features.Forthis,weemploy avarietyof measurestoevaluatethe classification basedon these features.Weshow thatby carefully collectingand synthesizing readily available background information, such as ratings for online reviews, one can accurately predict social links based on trust.

Keywords

Social Computing
Trust Prediction
Trust Inference
Epinions
Data Mining

Cited by (0)