Towards automatic filtering of fake reviews
NeurocomputingPublished 24 May 2018
Emerson F. Cardoso, Renato M. Silva, Tiago A. Almeida
Citations95
SJR quartileQ1
SJR score1.47
SNIP1.94
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
A comprehensive analysis of content-based classification methods for fake review detection using multiple settings, employing different types of learning and datasets provides sufficient evidence to respond appropriately to the open questions.
Abstract
Online opinions significantly influence consumer purchase decisions. Unfortunately, this has led to a dramatic increase of fake (or spam) reviews that can damage the reputation of brands and artificially manipulate users' perceptions about products and companies. Despite the efforts of several studies on fake review detection, important questions still remain open. For instance, there is no consensus if the performance of the classification methods is affected when they are used in real-world scenarios that require online learning. Moreover, it is also not known if the performance of the methods decreases due to the time-ordered nature of the reviews. To answer these and other important open questions, this work presents a comprehensive analysis of content-based classification methods for fake review detection. The experiments were performed in multiple settings, employing different types of learning and datasets. A careful analysis of the results provided sufficient evidence to respond appropriately to the open questions, which can be used as a baseline for future studies.
Keywords
Computer ScienceSocial Sciences
Machine LearningRandom Forests
126,192 Citations2001Leo Breiman
Internal estimates monitor error, strength, and correlation and these are used to show the response to increasing the number of features used in the forest, and are also applicable to regression.
Machine LearningSupport-vector networks
40,463 Citations1995Corinna Cortes, Vladimir Vapnik
High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated and the performance of the support- vector network is compared to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.
Machine LearningSupport-Vector Networks
33,035 Citations1995Corinna Cortes, Vladimir Vapnik
High generalization ability of support-vector networks utilizing polynomial input transformations is demonstrated and the performance of the support- vector network is compared to various classical learning algorithms that all took part in a benchmark study of Optical Character Recognition.
IEEE Transactions on Information TheoryNearest neighbor pattern classification
16,134 Citations1967Thomas M. Cover, Peter E. Hart
The nearest neighbor decision rule assigns to an unclassified sample point the classification of the nearest of a set of previously classified points, so it may be said that half the classification information in an infinite sample set is contained in the nearest neighbor.
Psychological ReviewThe perceptron: A probabilistic model for information storage and organization in the brain.
11,746 Citations1958Frank F. Rosenblatt
This article will be concerned primarily with the second and third questions, which are still subject to a vast amount of speculation, and where the few relevant facts currently supplied by neurophysiology have not yet been integrated into an acceptable theory.
Statistical Comparisons of Classifiers over Multiple Data Sets
11,198 Citations2006Janez Demšar
A set of simple, yet safe and robust non-parametric tests for statistical comparisons of classifiers is recommended: the Wilcoxon signed ranks test for comparison of two classifiers and the Friedman test with the corresponding post-hoc tests for comparisons of more classifiers over multiple data sets.
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.
Solving large scale linear prediction problems using stochastic gradient descent algorithms
1,150 Citations2004Tong Zhang
Stochastic gradient descent algorithms on regularized forms of linear prediction methods, related to online algorithms such as perceptron, are studied, and numerical rate of convergence for such algorithms is obtained.
Proceedings of the 15th ACM international conference on Information and knowledge management - CIKM '06
756 Citations2006Alfredo Cuzzocrea, James Allan, Norman W. Paton, Divesh Srivastava, Rakesh Agrawal, Andrei Z. Broder, Mohammed J. Zaki, K. Selçuk Candan, Alexandros Labrinidis, Assaf Schuster, Haixun Wang, Cuzzocrea, Alfredo Massimiliano +10 more
Spotting fake reviewer groups in consumer reviews
747 Citations2012Arjun Mukherjee, Bing Liu +1 more
This paper studies spam detection in the collaborative setting, i.e., to discover fake reviewer groups by using several behavioral models derived from the collusion phenomenon among fake reviewers and relation models based on the relationships among groups, individual reviewers, and products they reviewed to detectfake reviewer groups.
Proceedings of the International AAAI Conference on Web and Social MediaWhat Yelp Fake Review Filter Might Be Doing?
556 Citations2021Arjun Mukherjee, Vivek V. Venkataraman +2 more
This work attempts to find out what Yelp might be doing by analyzing its filtered reviews and postulates that Yelp’s filtering is reasonable and its filtering algorithm seems to be correlated with abnormal spamming behaviors.
Journal Of Big DataSurvey of review spam detection using machine learning techniques
495 Citations2015Mike Crawford, Taghi M. Khoshgoftaar +3 more
A strong and comprehensive comparative study of current research on detecting review spam using various machine learning techniques and to devise methodology for conducting further investigation is provided.
NeurocomputingA survey on data preprocessing for data stream mining: Current status and future directions
491 Citations2017Sergio Ramírez‐Gallego, Bartosz Krawczyk +3 more
This survey summarizes, categorize and analyze those contributions on data preprocessing that cope with streaming data, and takes into account the existing relationships between the different families of methods (feature and instance selection, and discretization).
Review spam detection
447 Citations2007Nitin Jindal, Bing Liu
This paper makes an attempt to study review spam and spam detection.
Spotting opinion spammers using behavioral footprints
417 Citations2013Arjun Mukherjee, Abhinav Kumar +5 more
An unsupervised model, called Author Spamicity Model (ASM), is proposed, which works in the Bayesian setting, which facilitates modeling spamicity of authors as latent and allows us to exploit various observed behavioral footprints of reviewers.
Proceedings of the International AAAI Conference on Web and Social MediaExploiting Burstiness in Reviews for Review Spammer Detection
354 Citations2021Geli Fei, Arjun Mukherjee +4 more
This work exploits the burstiness nature of reviews to identify review spammers and proposes a novel evaluation method to evaluate the detected spammers automatically using supervised classification of their reviews, which outperforms strong baselines.
NeurocomputingIncremental on-line learning: A review and comparison of state of the art algorithms
350 Citations2017Viktor Losing, Barbara Hammer +1 more
This work analyzes the key properties of eight popular incremental methods representing different algorithm classes and evaluates them with regards to their on-line classification error as well as to their behavior in the limit, facilitating the choice of the best method for a given application.
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.
Negative Deceptive Opinion Spam
291 Citations2013Myle Ott, Claire Cardie +1 more
This work creates and study the first dataset of deceptive opinion spam with negative sentiment reviews, and finds that standard n-gram text categorization techniques can detect negative deceptive opinions spam with performance far surpassing that of human judges.
Expert Systems with ApplicationsDetection of review spam: A survey
287 Citations2014Atefeh Heydari, Mohammad Ali Tavakoli +2 more
The present research focuses on systematically analyzing and categorizing models that detect review spam, and finds that studies can be categorized into three groups that focus on methods to detect spam reviews, individual spammers and group spam.
NeurocomputingDetecting spammers on social networks
253 Citations2015Xianghan Zheng, Zhipeng Zeng +3 more
A supervised machine learning based solution is proposed for an effective spammer detection and shows that the proposed solution is capable to provide excellent performance with true positive rate of spammers and non-spammers reaching 99.1% and 99.9% respectively.
Analyzing and Detecting Review Spam
225 Citations2007Nitin Jindal, Bing Liu
It is shown that review spam is quite different from Web page spam and email spam, and thus requires different detection techniques, and several techniques to detect them are proposed.
NeurocomputingSupport vector machines under adversarial label contamination
221 Citations2015Xiao Huang, Battista Biggio +4 more
This work considers an attacker that aims to maximize the SVM?s classification error by flipping a number of labels in the training data, and formalizes a corresponding optimal attack strategy, and solves it by means of heuristic approaches to keep the computational complexity tractable.
Comparison of Deceptive and Truthful Travel Reviews
217 Citations2009Kyung-Hyan Yoo, Ulrike Gretzel
Using findings from deception theory, a study was conducted to compare the language structure of deceptive and truthful hotel reviews, suggesting that it might be difficult to distinguish between deception and truthful reviews based on structural properties.
Information Processing & ManagementDetecting positive and negative deceptive opinions using PU-learning
174 Citations2014Donato Hernández Fusilier, Manuel Montes-y-Gómez +2 more
A novel method is proposed that with respect to its original version is much more conservative at the moment of selecting the negative examples from the unlabeled ones and consistently outperformed the original PU-learning approach in the detection of positive and negative deceptive opinions respectively.
NeurocomputingDocument representation and feature combination for deceptive spam review detection
150 Citations2017Luyang Li, Bing Qin +2 more
A neural network based model is presented to learn the representation of reviews and makes a hard attention through the composition from sentence representation into document representation, compute the importance weights of each sentence and incorporate them into the composition process of document representation.
Bimodal Distribution and Co-Bursting in Review Spam Detection
135 Citations2017Huayi Li, Geli Fei +5 more
A two-mode Labeled Hidden Markov Model is proposed to model spamming using only individual reviewers' review posting times and a co-bursting network based on co- Bursting relations is proposed, which helps detect groups of spammers more effectively than existing approaches.
Proceedings of the International AAAI Conference on Web and Social MediaAnalyzing and Detecting Opinion Spam on a Large-scale Dataset via Temporal and Spatial Patterns
119 Citations2021Huayi Li, Zhiyuan Chen +3 more
This work presents the first large-scale analysis of restaurant reviews filtered by Dianping's fake review filtering system, and proposes some novel temporal and spatial features for supervised opinion spam detection that significantly outperform existing state-of-art features.
IEEE Transactions on CyberneticsPrincipal Component 2-D Long Short-Term Memory for Font Recognition on Single Chinese Characters
102 Citations2015Dapeng Tao, Lin Xu +2 more
It is proposed that font recognition on a single Chinese character is a sequence classification problem, which can be effectively solved by recurrent neural networks and integrated a principal component convolution layer with the 2-D long short-term memory (2DLSTM) and developed principal component 2 DLSTM (PC-2DL STM) algorithm.
IEEE Transactions on Image ProcessingTensor Rank Preserving Discriminant Analysis for Facial Recognition
79 Citations2017Dapeng Tao, Yanan Guo +2 more
A new tensor-based feature extraction algorithm termed tensor rank preserving discriminant analysis (TRPDA) for facial image recognition is proposed; the proposed method involves two stages: in the first stage, the low-dimensional tensor subspace of the original input tensor samples was obtained and discriminative locality alignment was utilized to obtain the ultimate vector feature representation for subsequent facial recognition.
Procedia Computer ScienceThe Impact of applying Different Preprocessing Steps on Review Spam Detection
78 Citations2017Wael Etaiwi, Ghazi Al‐Naymat
This research will investigate the effects of preprocessing steps on the accuracy of reviews spam detection, and a labeled dataset of Hotels reviews will be analyze and process.
NeurocomputingImproved competitive learning neural networks for network intrusion and fraud detection
76 Citations2011John Zhong Lei, Ali A. Ghorbani
Experimental comparisons on both research data and real-world data in fraud detection and network intrusion detection demonstrate that both the ICLN and the SICLN achieve high performance, and the latter outperforms traditional unsupervised clustering algorithms.
NeurocomputingAn incremental extreme learning machine for online sequential learning problems
72 Citations2013Lu Guo, Jinghua Hao +1 more
The presented IELM algorithms have better performance than other incremental algorithms such as online sequential ELM (OS-ELM), a unified framework of LS-SVM and PSVM and fixed size LSSVM presented by Espinoza et al. (2006).
Detecting Deceptive Opinion Spam Using Human Computation
72 Citations2012Christopher G. Harris
This study pools deceptive reviews solicited through crowdsourcing with actual reviews obtained from product review websites to spot deceptive opinion spam and finds that the combination of human-based assessment methods with easily-obtained statistical information generated from the review text outperforms detection methods using human assessors alone.
Toward a Language Modeling Approach for Consumer Review Spam Detection
70 Citations2010Cheng-Yu Lai, Kaiqiang Xu +3 more
The experimental results confirm that the KL divergence and the probabilistic language modeling based computational model is effective for the detection of untruthful reviews and empower the development of a novel computational methodology to combat online review spam.
NeurocomputingSpam filtering for short messages in adversarial environment
69 Citations2014Patrick P. K. Chan, Cheng Yang +2 more
This paper proposes a good word attack strategy which maximizes the influence to a classifier with the least number of inserted characters based on the weight values and also the length of words and proposes the feature reweighting method with a new rescaling function which minimizes the importance of the feature representing a short word in order to require more inserted characters for a successful evasion.
NeurocomputingAn ensemble of classifiers for the diagnosis of erythemato-squamous diseases
62 Citations2006Loris Nanni
The results improved the average predictive accuracy obtained by a ''stand-alone'' SVM or by a RS ensemble of SVM.
Conceptual level similarity measure based review spam detection
58 Citations2010Siddu P. Algur, Amit Prakash Patil +2 more
An attempt to detect whether a review is a spam or a non spam review, in order to provide a trusted review to help the customer in taking the proper buying decision, and demonstrates the effectiveness of the proposed technique in detecting spam and non spam reviews.
Lecture notes in computer scienceTokenising, Stemming and Stopword Removal on Anti-spam Filtering Domain
55 Citations2006José R. Méndez, Eva Iglesias +3 more
NeurocomputingDiscovering social spammers from multiple views
55 Citations2016Hua Shen, Fenglong Ma +4 more
This work proposes a generalized social spammer detection framework by jointly integrating multiple view information and a novel social regularization term into a classification model, and introduces a simple strategy to fill the missing data for each view.
NeurocomputingCo-detecting social spammers and spam messages in microblogging via exploiting social contexts
44 Citations2016Fangzhao Wu, Jinyun Shu +2 more
The proposed unified approach utilizes the posting relations between users and messages to combine social spammer detection with spam message detection and introduces an efficient optimization algorithm to solve the model of the approach and proposes an accelerated method to tackle the most time-consuming step.
iSRD: Spam review detection with imbalanced data distributions
42 Citations2014Hamzah Al Najada, Xingquan Zhu
The method, iSRD, outperforms baseline methods for review spam detection and employs a bagging based approach to build a number of balanced datasets, through which it can train a set of spam classifiers and use their ensemble to detect review spams.
Engineering Applications of Artificial IntelligenceEvaluation of a set of new ORF kernel functions of SVM for speech recognition
28 Citations2013Xueying Zhang, Xiaofeng Liu +1 more
The Optimal Relaxation Factor (ORF) kernel function is presented, which is a set of new SVM kernel functions for speech recognition, and it is proved that the ORF function is a Mercer kernel function.
Expert Systems with ApplicationsTowards filtering undesired short text messages using an online learning approach with semantic indexing
27 Citations2017Renato M. Silva, Túlio C. Alberto +2 more
A new hybrid ensemble approach is proposed that combines the predictions obtained by the classifiers using the original text samples along with their variations created by applying text normalization and semantic indexing techniques, which can improve the text content quality and enhance the performance of the expert systems for spamming detection.
Knowledge-Based SystemsMDLText: An efficient and lightweight text classifier
27 Citations2016Renato M. Silva, Tiago A. Almeida +1 more
The proposed MDLText is an efficient, lightweight, scalable, and fast multinomial text classifier, which is based on the minimum description length principle, which exhibits fast incremental learning as well as being sufficiently robust to prevent overfitting, which are desirable features in real-world applications, large-scale problems, and online scenarios.
An Analysis of Machine Learning Methods for Spam Host Detection
25 Citations2012Renato M. Silva, Akebo Yamakami +1 more
A comprehensive performance evaluation of several established machine learning techniques used to automatically detect and filter hosts that disseminate web spam indicates that bagging of decision trees, multilayer perceptron neural networks, random forest and adaptive boosting of decision Trees are promising in the task of web spam classification.
NeurocomputingSpam filtering using semantic similarity approach and adaptive BPNN
24 Citations2012Cheng Hua Li, Jimmy Xiangji Huang
A new method that integrates three kinds of semantic similarity approaches for spam filtering as a case study of data mining application to construct a latent semantic feature space from training data with a statistical method.
Combating Comment Spam with Machine Learning Approaches
22 Citations2015Mansour Alsaleh, Abdulrahman Alarifi +2 more
This paper proposes and develops a comment spam detection mechanism that can be deployed as a browser plugin for inspecting the Document Object Model of the web page in question and remove comments with spam content.
Journal of Intelligent & Robotic SystemsPost or Block? Advances in Automatically Filtering Undesired Comments
12 Citations2014Túlio C. Alberto, Johannes V. Lochter +1 more
Experiments carried out with a real and public database indicate that support vector machines, logistic regression and stacking ensemble methods, trained with both attributes extracted from the text messages and posting information, are promising for the task of blocking undesired comments.
DergiPark (Istanbul University)Machine Learning Methods for Spamdexing Detection
11 Citations2013Tiago Almeida, Renato M. Silva +1 more
Results indicate that bagging of decision trees, multilayer perceptron neural networks, random forest and adaptive boosting of decision Trees are promising in the task of web spam classification.
NeurocomputingWebspam demotion: Low complexity node aggregation methods
7 Citations2011Thomas Largillier, Sylvain Peyronnet
This paper proposes to study the effects of node aggregation on the well-known ranking algorithm of Google (the PageRank) in the presence of Webspam, and presents four lightweight aggregation techniques suitable for its size.