A Content-Aware Trust Index for Online Review Spam Detection - Data and Applications Security and Privacy XXXI
Conference Papers Year : 2017

A Content-Aware Trust Index for Online Review Spam Detection

Hao Xue
  • Function : Author
  • PersonId : 1026642
Fengjun Li
  • Function : Author
  • PersonId : 1026643

Abstract

Online review helps reducing uncertainty in the pre-purchasing decision phase and thus becomes an important information source for consumers. With the increasing popularity of online review systems, a large volume of reviews of varying quality is generated. Meanwhile, individual and professional spamming activities have been observed in almost all online review platforms. Deceptive reviews with fake ratings or fake content are inserted into the system to influence people’s perception from reading these reviews. The deceptive reviews and reviews of poor quality significantly affect the effectiveness of online review systems. In this work, we define novel aspect-specific indicators that measure the deviations of aspect-specific opinions of a review from the aggregated opinions. Then, we propose a three-layer trust framework that relies on aspect-specific indicators to ascertain veracity of reviews and compute trust scores of their reviewers. An iterative algorithm is developed for propagation of trust scores in the three-layer trust framework. The converged trust score of a reviewer is a credibility indicators that reflects the trustworthiness of the reviewer and the quality of his reviews, which becomes an effective trust index for online review spam detection.
Fichier principal
Vignette du fichier
453481_1_En_27_Chapter.pdf (475.06 Ko) Télécharger le fichier
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-01684367 , version 1 (15-01-2018)

Licence

Identifiers

Cite

Hao Xue, Fengjun Li. A Content-Aware Trust Index for Online Review Spam Detection. 31th IFIP Annual Conference on Data and Applications Security and Privacy (DBSEC), Jul 2017, Philadelphia, PA, United States. pp.489-508, ⟨10.1007/978-3-319-61176-1_27⟩. ⟨hal-01684367⟩
189 View
151 Download

Altmetric

Share

More