Detecting opinion spammer groups and spam targets through community discovery and sentiment analysis

Euijin Choo, Ting Yu, Min Chi

Research output: Contribution to journalReview article

2 Citations (Scopus)

Abstract

In this paper we investigate on detecting opinion spammer groups through analyzing how users interact with each other. More specifically, our approaches are based on 1) discovering strong vs. weak implicit communities by mining user interaction patterns, and 2) revealing positive vs. negative communities through sentiment analysis on user interactions. Through extensive experiments over various datasets collected from Amazon, we found that the discovered strong, positive communities are significantly more likely to be opinion spammer groups than other communities. Interestingly, while our approach focused mainly on the characteristics of user interactions, it is comparable to the state of the art content-based classifier that mainly uses various content-based features extracted from user reviews. More importantly, we argue that our approach can be more robust than the latter in that if spammers superficially alter their review contents, our approach can still reliably identify them while the content-based approaches may fail.

Original languageEnglish
Pages (from-to)283-318
Number of pages36
JournalJournal of Computer Security
Volume25
Issue number3
DOIs
Publication statusPublished - 1 Jan 2017

Fingerprint

Classifiers
Experiments

Keywords

  • community discovery
  • Opinion spammer groups
  • sentiment analysis

ASJC Scopus subject areas

  • Software
  • Safety, Risk, Reliability and Quality
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

Detecting opinion spammer groups and spam targets through community discovery and sentiment analysis. / Choo, Euijin; Yu, Ting; Chi, Min.

In: Journal of Computer Security, Vol. 25, No. 3, 01.01.2017, p. 283-318.

Research output: Contribution to journalReview article

@article{730184e6b8574eb88e71f2da4542d778,
title = "Detecting opinion spammer groups and spam targets through community discovery and sentiment analysis",
abstract = "In this paper we investigate on detecting opinion spammer groups through analyzing how users interact with each other. More specifically, our approaches are based on 1) discovering strong vs. weak implicit communities by mining user interaction patterns, and 2) revealing positive vs. negative communities through sentiment analysis on user interactions. Through extensive experiments over various datasets collected from Amazon, we found that the discovered strong, positive communities are significantly more likely to be opinion spammer groups than other communities. Interestingly, while our approach focused mainly on the characteristics of user interactions, it is comparable to the state of the art content-based classifier that mainly uses various content-based features extracted from user reviews. More importantly, we argue that our approach can be more robust than the latter in that if spammers superficially alter their review contents, our approach can still reliably identify them while the content-based approaches may fail.",
keywords = "community discovery, Opinion spammer groups, sentiment analysis",
author = "Euijin Choo and Ting Yu and Min Chi",
year = "2017",
month = "1",
day = "1",
doi = "10.3233/JCS-16941",
language = "English",
volume = "25",
pages = "283--318",
journal = "Journal of Computer Security",
issn = "0926-227X",
publisher = "IOS Press",
number = "3",

}

TY - JOUR

T1 - Detecting opinion spammer groups and spam targets through community discovery and sentiment analysis

AU - Choo, Euijin

AU - Yu, Ting

AU - Chi, Min

PY - 2017/1/1

Y1 - 2017/1/1

N2 - In this paper we investigate on detecting opinion spammer groups through analyzing how users interact with each other. More specifically, our approaches are based on 1) discovering strong vs. weak implicit communities by mining user interaction patterns, and 2) revealing positive vs. negative communities through sentiment analysis on user interactions. Through extensive experiments over various datasets collected from Amazon, we found that the discovered strong, positive communities are significantly more likely to be opinion spammer groups than other communities. Interestingly, while our approach focused mainly on the characteristics of user interactions, it is comparable to the state of the art content-based classifier that mainly uses various content-based features extracted from user reviews. More importantly, we argue that our approach can be more robust than the latter in that if spammers superficially alter their review contents, our approach can still reliably identify them while the content-based approaches may fail.

AB - In this paper we investigate on detecting opinion spammer groups through analyzing how users interact with each other. More specifically, our approaches are based on 1) discovering strong vs. weak implicit communities by mining user interaction patterns, and 2) revealing positive vs. negative communities through sentiment analysis on user interactions. Through extensive experiments over various datasets collected from Amazon, we found that the discovered strong, positive communities are significantly more likely to be opinion spammer groups than other communities. Interestingly, while our approach focused mainly on the characteristics of user interactions, it is comparable to the state of the art content-based classifier that mainly uses various content-based features extracted from user reviews. More importantly, we argue that our approach can be more robust than the latter in that if spammers superficially alter their review contents, our approach can still reliably identify them while the content-based approaches may fail.

KW - community discovery

KW - Opinion spammer groups

KW - sentiment analysis

UR - http://www.scopus.com/inward/record.url?scp=85020095578&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85020095578&partnerID=8YFLogxK

U2 - 10.3233/JCS-16941

DO - 10.3233/JCS-16941

M3 - Review article

AN - SCOPUS:85020095578

VL - 25

SP - 283

EP - 318

JO - Journal of Computer Security

JF - Journal of Computer Security

SN - 0926-227X

IS - 3

ER -