Arjun Mukherjee

Spotting fake reviewer groups in consumer reviews

Opinionated social media such as product reviews are now widely used by individuals and organizations for their decision making. However, due to the reason of profit or fame, people try to game the system by opinion spamming (e.g., writing fake reviews) to promote or demote some target products. For reviews to reflect genuine user experiences and opinions, such spam reviews should be detected. Prior works on opinion spam focused on detecting fake reviews and individual fake reviewers. However, a fake reviewer group (a group of reviewers who work collaboratively to write fake reviews) is even more damaging as they can take total control of the sentiment on the target product due to its size. This paper studies spam detection in the collaborative setting, i.e., to discover fake reviewer groups. The proposed method first uses a frequent itemset mining method to find a set of candidate groups. It then uses 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 detect fake reviewer groups. Additionally, we also built a labeled dataset of fake reviewer groups. Although labeling individual fake reviews and reviewers is very hard, to our surprise labeling fake reviewer groups is much easier. We also note that the proposed technique departs from the traditional supervised learning approach for spam detection because of the inherent nature of our problem which makes the classic supervised learning approach less effective. Experimental results show that the proposed method outperforms multiple strong baselines including the state-of-the-art supervised classification, regression, and learning to rank algorithms.

Spotting opinion spammers using behavioral footprints

Opinionated social media such as product reviews are now widely used by individuals and organizations for their decision making. However, due to the reason of profit or fame, people try to game the system by opinion spamming (e.g., writing fake reviews) to promote or to demote some target products. In recent years, fake review detection has attracted significant attention from both the business and research communities. However, due to the difficulty of human labeling needed for supervised learning and evaluation, the problem remains to be highly challenging. This work proposes a novel angle to the problem by modeling spamicity as latent. An unsupervised model, called Author Spamicity Model (ASM), is proposed. It works in the Bayesian setting, which facilitates modeling spamicity of authors as latent and allows us to exploit various observed behavioral footprints of reviewers. The intuition is that opinion spammers have different behavioral distributions than non-spammers. This creates a distributional divergence between the latent population distributions of two clusters: spammers and non-spammers. Model inference results in learning the population distributions of the two clusters. Several extensions of ASM are also considered leveraging from different priors. Experiments on a real-life Amazon review dataset demonstrate the effectiveness of the proposed models which significantly outperform the state-of-the-art competitors.

Detecting group review spam

It is well-known that many online reviews are not written by genuine users of products, but by spammers who write fake reviews to promote or demote some target products. Although some existing works have been done to detect fake reviews and individual spammers, to our knowledge, no work has been done on detecting spammer groups. This paper focuses on this task and proposes an effective technique to detect such groups.

Aspect Extraction with Automated Prior Knowledge Learning

Aspect extraction is an important task in sentiment analysis. Topic modeling is a popular method for the task. However, unsupervised topic models often generate incoherent aspects. To address the issue, several knowledge-based models have been proposed to incorporate prior knowledge provided by the user to guide modeling. In this paper, we take a major step forward and show that in the big data era, without any user input, it is possible to learn prior knowledge automatically from a large amount of review data available on the Web. Such knowledge can then be used by a topic model to discover more coherent aspects. There are two key challenges: (1) learning quality knowledge from reviews of diverse domains, and (2) making the model fault-tolerant to handle possibly wrong knowledge. A novel approach is proposed to solve these problems. Experimental results using reviews from 36 domains show that the proposed approach achieves significant improvements over state-of-the-art baselines.

Discovering coherent topics using general knowledge

Topic models have been widely used to discover latent topics in text documents. However, they may produce topics that are not interpretable for an application. Researchers have proposed to incorporate prior domain knowledge into topic models to help produce coherent topics. The knowledge used in existing models is typically domain dependent and assumed to be correct. However, one key weakness of this knowledge-based approach is that it requires the user to know the domain very well and to be able to provide knowledge suitable for the domain, which is not always the case because in most real-life applications, the user wants to find what they do not know. In this paper, we propose a framework to leverage the general knowledge in topic models. Such knowledge is domain independent. Specifically, we use one form of general knowledge, i.e., lexical semantic relations of words such as synonyms, antonyms and adjective attributes, to help produce more coherent topics. However, there is a major obstacle, i.e., a word can have multiple meanings/senses and each meaning often has a different set of synonyms and antonyms. Not every meaning is suitable or correct for a domain. Wrong knowledge can result in poor quality topics. To deal with wrong knowledge, we propose a new model, called GK-LDA, which is able to effectively exploit the knowledge of lexical relations in dictionaries. To the best of our knowledge, GK-LDA is the first such model that can incorporate the domain independent knowledge. Our experiments using online product reviews show that GK-LDA performs significantly better than existing state-of-the-art models.

Mining contentions from discussions and debates

Social media has become a major source of information for many applications. Numerous techniques have been proposed to analyze network structures and text contents. In this paper, we focus on fine-grained mining of contentions in discussion/debate forums. Contentions are perhaps the most important feature of forums that discuss social, political and religious issues. Our goal is to discover contention and agreement indicator expressions, and contention points or topics both at the discussion collection level and also at each individual post level. To the best of our knowledge, limited work has been done on such detailed analysis. This paper proposes three models to solve the problem, which not only model both contention/agreement expressions and discussion topics, but also, more importantly, model the intrinsic nature of discussions/debates, i.e., interactions among discussants or debaters and topic sharing among posts through quoting and replying relations. Evaluation results using real-life discussion/debate posts from several domains demonstrate the effectiveness of the proposed models.

Detecting Campaign Promoters on Twitter Using Markov Random Fields

As social media is becoming an increasingly important source of public information, companies, organizations and individuals are actively using social media platforms to promote their products, services, ideas and ideologies. Unlike promotional campaigns on TV or other traditional mass media platforms, campaigns on social media often appear in stealth modes. Campaign promoters often try to influence peoples behaviors/opinions/decisions in a latent manner such that the readers are not aware that the messages they see are strategic campaign posts aimed at persuading them to buy target products/services. Readers take such campaign posts as just organic posts from the general public. It is thus important to discover such campaigns, their promoter accounts and how the campaigns are organized and executed as it can uncover the dynamics of Internet marketing. This discovery is clearly useful for competitors and also the general public. However, so far little work has been done to solve this problem. In this paper, we study this important problem in the context of the Twitter platform. Given a set of tweets streamed from Twitter based on a set of keywords representing a particular topic, the proposed technique aims to identify user accounts that are involved in promotion. We formulate the problem as a relational classification problem and solve it using typed Markov Random Fields (T-MRF), which is proposed as a generalization of the classic Markov Random Fields. Our experiments are carried out using three real-life datasets from the health science domain related to smoking. Such campaigns are interesting to health scientists, government health agencies and related businesses for obvious reasons. Our results show that the proposed method is highly effective.

On the Temporal Dynamics of Opinion Spamming: Case Studies on Yelp

Recently, the problem of opinion spam has been widespread and has attracted a lot of research attention. While the problem has been approached on a variety of dimensions, the temporal dynamics in which opinion spamming operates is unclear. Are there specific spamming policies that spammers employ? What kind of changes happen with respect to the dynamics to the truthful ratings on entities. How do buffered spamming operate for entities that need spamming to retain threshold popularity and reduced spamming for entities making better success? We analyze these questions in the light of time-series analysis on Yelp. Our analyses discover various temporal patterns and their relationships with the rate at which fake reviews are posted. Building on our analyses, we employ vector autoregression to predict the rate of deception across different spamming policies. Next, we explore the effect of filtered reviews on (long-term and imminent) future rating and popularity prediction of entities. Our results discover novel temporal dynamics of spamming which are intuitive, arguable and also render confidence on Yelps filtering. Lastly, we leverage our discovered temporal patterns in deception detection. Experimental results on large-scale reviews show the effectiveness of our approach that significantly improves the existing approaches.

Detecting Deceptive Opinion Spam using Linguistics, Behavioral and Statistical Modeling

With the advent of Web 2.0, consumer reviews have become an important resource for public opinion that influence our decisions over an extremely wide spectrum of daily and professional activities: e.g., where to eat, where to stay, which products to purchase, which doctors to see, which books to read, which universities to attend, and so on. Positive/negative reviews directly translate to financial gains/losses for companies. This unfortunately gives strong incentives for opinion spamming which refers to illegal human activities (e.g., writing fake reviews and giving false ratings) that try to mislead customers by promoting/demoting certain entities (e.g., products and businesses). The problem has been widely reported in the news. Despite the recent research efforts on detection, the problem is far from solved. What is worse is that opinion spamming is widespread. While credit card fraud is as rare as 0.2%, based on our research we estimated that up to 30% of the reviews on many Web sites could be fake. Thus, detecting fake reviews and opinions is a pressing and also profound issue as it is critical to ensure the trustworthiness of the information on the web. Without detecting them, the social media could become a place full of lies, fakes, and deceptions, and completely useless. Major review hosting sites and e-commerce vendors have already made some progress in detecting fake reviews. However, the task is still extremely challenging because it is very difficult to obtain large-scale ground truth samples of deceptive opinions for algorithm development and for evaluation, or to conduct large-scale domain expert evaluations. Further, in contrast to other kinds of spamming (e.g., Web and link spam, social/blog spam, email spam, etc.) opinion spam has a very unique flavor as it involves fluid sentiments of users and their evaluations. Thus, they require a very different treatment. Since our first paper in 2007 (Jindal and Liu, 2007) on the topic, our group and many other researchers have proposed several algorithms and bridged algorithmic methodologies from various scientific disciplines including computational linguistics (Ott et al., 2011), social and behavioral sciences (Jindal and Liu, 2008; Mukherjee et al., 2013a, b), machine learning, data mining and Bayesian statistics (Mukherjee et al., 2012; Fei et al., 2013; Mukherjee et al., 2013c; Li et al., 2014b; Li et al., 2014a) to solve the problem. The field of deceptive opinion spam has gained a lot of interest in communications (Hancock et al., 2008), psycholinguistics communities (Gokhman et al., 2012), and economic analysis (Wang, 2010) apart from mainstream NLP and Web mining as attested by publications in top tier venues in their respective communities. The problem has far reaching implications in various allied NLP topics including Lie Detection, Forensic Linguistics, Opinion Trust and Veracity Verification and Plagiarism Detection. However, owing to the inherent nature of the problem, a unique blend of NLP, data mining, machine learning, social, behavioral, and statistical techniques are required which many NLP researchers may not be familiar with. In this tutorial, we aim to cover the problem in its full depth and width, covering diverse algorithms that have been developed over the past 7 years. The most attractive quality of these techniques is that many of them can be adapted for cross-domain and unsupervised settings. Some of the methods are even in use by startups and established companies. Our focus is on insight and understanding, using illustrations and intuitive deductions. The goal of the tutorial is to make the inner workings of these techniques transparent, intuitive and their results interpretable.

Analysis of Anxious Word Usage on Online Health Forums.

Online health communities and support groups are a valuable source of information for users suffering from a physical or mental illness. Users turn to these forums for moral support or advice on specific conditions, symptoms, or side effects of medications. This paper describes and studies the linguistic patterns of a community of support forum users over time focused on the used of anxious related words. We introduce a methodology to identify groups of individuals exhibiting linguistic patterns associated with anxiety and the correlations between this linguistic pattern and other word usage. We find some evidence that participation in these groups does yield positive effects on their users by reducing the frequency of anxious related word used over time.