Using PU-Learning to Detect Deceptive Opinion Spam
Published 1 June 2013
Donato Hernández, Rafael Guzmán-Cabrera, Manuel Montes y Gómez, Paolo Rosso
Citations55
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
This paper focuses on the detection of deceptive opinion spam, which consists of fictitious opinions that have been deliberately written to sound authentic, in order to deceive the consumers and proposes a method based on the PU-learning approach which learns only from a few positive examples and a set of unlabeled data.
Abstract
Nowadays a large number of opinion reviews are posted on the Web. Such reviews are a very important source of information for customers and companies. The former rely more than ever on online reviews to make their purchase decisions and the latter to respond promptly to their clients’ expectations. Due to the economic importance of these reviews there is a growing trend to incorporate spam on such sites, and, as a consequence, to develop methods for opinion spam detection. In this paper we focus on the detection of deceptive opinion spam, which consists of fictitious opinions that have been deliberately written to sound authentic, in order to deceive the consumers. In particular we propose a method based on the PU-learning approach which learns only from a few positive examples and a set of unlabeled data. Evaluation results in a corpus of hotel reviews demonstrate the appropriateness of the proposed method for real applications since it reached a f-measure of 0.84 in the detection of deceptive opinions using only 100 positive examples for training.
Keywords
Computer ScienceSocial Sciences
ACM SIGKDD Explorations NewsletterThe WEKA data mining software
17,849 Citations2009Mark Hall, Eibe Frank +4 more
This paper provides an introduction to the WEKA workbench, reviews the history of the project, and, in light of the recent 3.6 stable release, briefly discusses what has been added since the last stable version (Weka 3.4) released in 2003.
Encyclopedia of Machine Learning and Data MiningSentiment Analysis and Opinion Mining
3,319 Citations2017Lei Zhang, Bing Liu
Opinion spam and analysis
1,520 Citations2008Nitin Jindal, Bing Liu
It is shown that opinion spam is quite different from Web spam and email spam, and thus requires different detection techniques, and therefore requires some novel techniques to detect them.
IEEE Transactions on Neural NetworksSupport vector machines for spam categorization
1,457 Citations1999Harris Drucker, Donghui Wu +1 more
The use of support vector machines in classifying e-mail as spam or nonspam is studied by comparing it to three other classification algorithms: Ripper, Rocchio, and boosting decision trees, which found SVM's performed best when using binary features.
One-class svms for document classification
1,225 Citations2002Larry M. Manevitz, Malik Yousef
The SVM approach as represented by Schoelkopf was superior to all the methods except the neural network one, where it was, although occasionally worse, essentially comparable.
Detecting product review spammers using rating behaviors
720 Citations2010Ee‐Peng Lim, Viet-An Nguyen +3 more
This paper identifies several characteristic behaviors of review spammers and model these behaviors so as to detect the spammers, and shows that the detected spammers have more significant impact on ratings compared with the unhelpful reviewers.
arXiv (Cornell University)Finding Deceptive Opinion Spam by Any Stretch of the Imagination
688 Citations2011Myle Ott, Yejin Choi +2 more
Building text classifiers using positive and unlabeled examples
672 Citations2004B. Liu, Y. Dai +3 more
A more principled approach to solving the problem of building text classifiers using positive and unlabeled examples based on a biased formulation of SVM is proposed, and it is shown experimentally that it is more accurate than the existing techniques.
Detecting spam web pages through content analysis
612 Citations2006Alexandros Ntoulas, Marc Najork +2 more
Some previously-undescribed techniques for automatically detecting spam pages are considered, and the effectiveness of these techniques in isolation and when aggregated using classification algorithms is examined.
Partially Supervised Classification of Text Documents
516 Citations2002Bing Liu, Wee Sun Lee +2 more
This paper shows that the problem of identifying documents from a set of documents of a particular topic or class P and a large set M of mixed documents, and that under appropriate conditions, solutions to the constrained optimization problem will give good solution to the partially supervised classification problem.
Finding unusual review patterns using unexpected rules
339 Citations2010Nitin Jindal, Bing Liu +1 more
Using the technique, an Amazon.com review dataset is analyzed and many unexpected rules and rule groups which indicate spam activities are found.
The lie detector
339 Citations2009Rada Mihalcea, Carlo Strapparava
It is shown that automatic classification is a viable technique to distinguish between truth and falsehood as expressed in language and a method for class-based feature analysis is introduced, which sheds some light on the features that are characteristic for deceptive text.
Detecting group review spam
235 Citations2011Arjun Mukherjee, Bing Liu +3 more
This paper proposes an effective technique to detect spammer groups and focuses on the task of detecting fake reviews written by spammers.
Decision Support SystemsMaking objective decisions from subjective data: Detecting irony in customer reviews
171 Citations2012Antonio Reyes, Paolo Rosso
A set of customer reviews, which are considered ironic both in social and mass media, are analyzed to gather a set of discriminating elements to represent irony, in particular, the kind of irony expressed in such reviews.
IEEE Transactions on Knowledge and Data EngineeringA Statistical Language Modeling Approach to Online Deception Detection
101 Citations2008Lina Zhou, Yongmei Shi +1 more
The results showed that the proposed SLM approach to deception detection outperformed a state-of-the-art text categorization method as well as traditional feature-based methods.
Merging multiple criteria to identify suspicious reviews
63 Citations2010Guangyu Wu, Derek Greene +1 more
A number of criteria that might be indicative of suspicious reviews are presented and alternative methods for integrating these criteria to produce a unified "suspiciousness" ranking are evaluated.
Review spam detection via time series pattern discovery
56 Citations2012Sihong Xie, Guan Wang +2 more
It is discovered that singleton review is a significant source of spam reviews and largely affects the ratings of online stores, and the proposed method to detect such attacks via unusually correlated temporal patterns is effective.
Journal of ComputersReliable Negative Extracting Based on kNN for Learning from Positive and Unlabeled Examples
42 Citations2009Bangzuo Zhang, Wanli Zuo
A new reliable negative extracting algorithm for step 1 is proposed that adopts kNN algorithm to rank the similarity of unlabeled examples from the k nearest positive examples, and set a threshold to label some unlabeling examples that lower than it as the reliable negative examples rather than the common method to label positive examples.
RiuNet (Universitat Politècnica de València)On the Use of PU Learning for Quality Flaw Prediction in Wikipedia.
10 Citations2012Edgardo Ferretti, Donato Hernández Fusilier +4 more
The work of Paolo Rosso was carried out also in the framework of the MICINN Text-Enterprise research project and the Microcluster VLC/Campus (International Campus of Excellence) on Multimodal Intelligent Systems.