Technical Merits:

Given the critical issues of opinion fraud in online communities, how can one identify fake reviews and attribute responsible culprits behind them? By conjoining expertise of the PIs over various modalities of deception footprints ranging over language, user behavior, and relational information, this project presents a research program that will result in much needed solutions to this emergent, prevalent, and socially impactful problem. The ultimate goal is to create a unified detection framework via synergistic integration of multiple information sources; from linguistics, user behavior, and network effects, to obtain the best of all worlds. The main idea is to formulate the problem as a relational inference task on composite heterogeneous networks, providing a principled, extensible approach that can blend and reinforce all the above cues towards effective and robust detection of fraud. From a scientific point of view, the research brings together three disciplines: natural language analysis, behavioral modeling, and graph mining. The outcome is a suite of novel, principled, and scalable techniques and models that will enhance our understanding of the creation and dissemination of opinion fraud and misinformation in general at a large scale. The PIs will collaborate with industry partners such as Yelp, Google, and Amazon, directly solicit online fake reviews, and conduct well-designed user studies for testing and validation of their techniques.

Broader Impacts:

The broader impact of our work is that it will enable the development of opinion fraud and misinformation detection solutions that are critical in achieving integrity and credibility on the Web. The outcome of this research will be beneficial to billions of Web users, governments, law enforcement agencies, multi-billion-dollar industries and service providers. As such, the two main bodies that this project will directly and significantly impact are the Web users and the e-commerce site owners. The PIs will collaborate with Yelp in evaluation and integration of their developed techniques and tools. The PIs will further reach out to other industry contacts at Amazon, Google, and TripAdvisor and aim to disseminate research results to them through published manuscripts and tutorials at major conferences where many industry practitioners attend, as well as release publicly available open-source software for opinion fraud detection. The public will also be educated through reaching out to popular press media for interviews and educational press articles.

3.2. Results

3.3. Related Publications

• PairNorm: Tackling Oversmoothing in GNNs
  Lingxiao Zhao  &  Leman Akoglu
  ICLR 2020, Addis Ababa, Ethiopia  
  


• Higher-Order Label Homogeneity and Spreading in Graphs
  Dhivya Eswaran, Srijan Kumar, Christos Faloutsos
  The WebConf (formerly WWW), Taipei, Taiwan, April 2020.
  
  
• Learning On-The-Job to Re-rank Anomalies from Top-1 Feedback
  Hemank Lamba  &  Leman Akoglu
  SIAM SDM 2019, Calgary, Canada  
  Best Research Paper Award    


• Fast and Accurate Anomaly Detection in Dynamic Graphs with a Two-Pronged Approach,
  Minji Yoon, Bryan Hooi, Kijung Shin, Christos Faloutsos,
  KDD ’19, August 4–8, 2019, Anchorage, AK, USA
  
  
• SedanSpot: Detecting Anomalies in Edge Streams
  Dhivya Eswaran, Christos Faloutsos
  IEEE ICDM, Singapore, November 2018.
  
  
• Social-Affiliation Networks: Patterns and the SOAR Model
  Dhivya Eswaran, Reihaneh Rabbany, Artur Dubrawski, Christos Faloutsos
  ECML PKDD, Dublin, Ireland, September 2018.
  
  
• A Quest for Structure: Jointly Learning the Graph Structure and Semi-Supervised Classification
  Xuan Wu^∗, Lingxiao Zhao^∗, Leman Akoglu
  ACM CIKM 2018, Turin, Italy  
  


• Incorporating Privileged Information to Unsupervised Anomaly Detection
  Shubhranshu Shekhar  &  Leman Akoglu
  ECML PKDD 2018, Dublin, Ireland  
  Best Student Machine Learning Paper Runner-up Award


• Beyond Outlier Detection: LookOut for Pictorial Explanation
  Nikhil Gupta, Dhivya Eswaran, Neil Shah, Leman Akoglu, Christos Faloutsos
  ECML PKDD 2018, Dublin, Ireland  
  
  
• Explaining Anomalies in Groups with Characterizing Subspace Rules
  Meghanath Macha and Leman Akoglu
  ECML PKDD 2018 (Journal Track), Dublin, Ireland  
  
  
• Contextual Outlier Detection with Multiple Contexts: Application to Ad Fraud
  Meghanath Macha, Deepak Pai, Leman Akoglu
  ECML PKDD 2018, Dublin, Ireland  
  
  
• Robust Semi-Supervised Learning on Multiple Networks with Noise
  Junting Ye and Leman Akoglu
  PAKDD 2018, Melbourne, Australia  
  
  
• ZooBP: Belief Propagation for Heterogeneous Networks
  Dhivya Eswaran, Stephan Gunnemann, Christos Faloutsos, Disha Makhija, Mohit Kumar
  VLDB, Munich, Germany, August 2017.  
  
  
• The Power of Certainty: A Dirichlet-Multinomial Model for Belief Propagation
  Dhivya Eswaran, Stephan Gunnemann, Christos Faloutsos
  SIAM SDM, Houston, USA, April 2017.  
  
  
• Bimodal Distribution and Co-Bursting in Review Spam Detection  
  Huayi Li, Geli Fei, Shuai Wang, Bing Liu, Weixiang Shao, Arjun Mukherjee and Jidong Shao.
  Proceedings of International World Wide Web Conference (WWW-2017), April 3-7, 2017, Perth, Australia.  
  
  
• FRAUDAR: Bounding Graph Fraud in the Face of Camouflage.  
  Bryan Hooi, Hyun Ah Song, Alex Beutel, Neil Shah, Kijung Shin, and Christos Faloutsos.
  KDD 2016, KDD Best Paper Award (Research Track). San Francisco, CA.  
  
  
• Sequential Ensemble Learning for Outlier Detection: A Bias-Variance Perspective  
  [arXiv]   [project page]   [ODDS data repo]  
  Shebuti Rayana, Wen Zhong, Leman Akoglu
  IEEE ICDM 2016, Barcelona, Spain 
  
  
• EdgeCentric: Anomaly Detection in Edge-Attributed Networks   [code]
  Neil Shah, Alex Beutel, Bryan Hooi, Leman Akoglu, Stephan Gunnemann, Disha Makhija, Mohit Kumar, Christos Faloutsos.
  IEEE ICDM Workshop on Data Mining for Cyber Security, Barcelona, Spain, Dec. 2016.  
  
  
• Less is More: Building Selective Anomaly Ensembles  
  Shebuti Rayana and Leman Akoglu
  Transactions on Knowledge Discovery from Data (TKDD), 10(4): 42, 2016.  
  
  
• Mining Big Time-series Data on the Web
  Yasushi Sakurai, Yasuko Matsubara, and Christos Faloutsos
  Proceedings of the 25th International Conference Companion on World Wide Web, WWW '16 Companion, pp 1029-1032, 2016.  
  
  
• Non-Linear Mining of Competing Local Activities
  Yasuko Matsubara, Yasushi Sakurai, and Christos Faloutsos
  Proceedings of the 25th International Conference on World Wide Web, WWW '16, 2016.  
  
  
• Temporal Opinion Spam Detection by Multivariate Indicative Signals
  Junting Ye, Santhosh Kumar, Leman Akoglu
  ICWSM, Cologne, Germany, May 2016.  
  
  
• GOTCHA! Network-based Fraud Detection for Social Security Fraud  
  Veronique Van Vlasselaer, Tina Eliassi-Rad, Leman Akoglu, Monique Snoeck, Bart Baesens
  Management Science (INFORMS), MS-14-00232.R4, 2016.  
  
  
• Collective Opinion Spam Detection using Active Inference  
  Shebuti Rayana and Leman Akoglu
  SIAM SDM, Miami, Florida, May 2016.  
  
  
• BIRDNEST: Bayesian Inference for Ratings-Fraud Detection  
  Bryan Hooi, Neil Shah, Alex Beutel, Stephan Gunnemann, Leman Akoglu, Mohit Kumar, Disha Makhija, Christos Faloutsos
  SIAM SDM, Miami, Florida, May 2016.  
  
  
• Catching Synchronized Behaviors in Large Networks: A Graph Mining Approach.  
  Meng Jiang, Peng Cui, Alex Beutel, Christos Faloutsos and Shiqiang Yang.
  ACM Transactions on Knowledge Discovery from Data (TKDD), 2015. (Best papers in KDD 2014, Special Issue - to appear).
  
  
• Analyzing and Detecting Opinion Spam on a Large-scale Dataset via Temporal and Spatial Patterns.  
  Huayi Li, Zhiyuan Chen, Arjun Mukherjee, Bing Liu, Jidong Shao.
  International AAAI Conference on Web and Social Media (ICWSM), Oxford, UK, May 2015.
  
  
• Discovering Opinion Spammer Groups by Network Footprints.  
  Junting Ye and Leman Akoglu.
  ECML/PKDD, Porto, Portugal, Sep. 2015.
  
  
• Collective Opinion Spam Detection: Bridging Review Networks and Metadata  
  Shebuti Rayana and Leman Akoglu
  ACM SIGKDD, Sydney, AU, Aug. 2015.
  
  
• RSC: Mining and Modeling Temporal Activity in Social Media.  
  Alceu Ferraz Costa, Yuto Yamaguchi, Agma Juci Machado Traina, Caetano Traina Jr. and Christos Faloutsos.
  KDD, Sydney, Australia, Aug. 2015.
  
  
• ACCAMS: Additive Co-Clustering to Approximate Matrices Succinctly.  
  Alex Beutel, Amr Ahmed, Alexander Smola.
  24th International World Wide Web Conference (WWW), Florence, Italy, May 2015.
  
  
• The Web as a Jungle: Non-Linear Dynamical Systems for Co-evolving Online Activities.  
  Yasuko Matsubara, Yasushi Sakurai, Christos Faloutsos.
  24th International World Wide Web Conference (WWW), Florence, Italy, May 2015.
  
  
• Event Detection and Factuality Assessment with Non-Expert Supervision.  
  Kenton Lee, Yoav Artzi, Yejin Choi, Luke Zettlemoyer.
  Conference on Empirical Methods in Natural Language Processing (EMNLP), 2015.
  
  
• Keystroke Patterns as Prosody in Digital Writings: A Case Study with Deceptive Reviews and Essays.  
  Song Feng, Ritwik Banerjee, Jun Seok Kang and Yejin Choi.
  Conference on Empirical Methods in Natural Language Processing (EMNLP), 2014.
  
  
• Detecting Campaign Promoters on Twitter using Markov Random Fields  
  Huayi Li, Arjun Mukherjee, Bing Liu, Rachel Kornfield, Sherry L. Emery
  IEEE International Conference on Data Mining (ICDM'14), December 14-17, 2014 - Shenzhen, China.
  
  
• Spotting Fake Reviews via Collective Positive-Unlabeled Learning  
  Huayi Li, Zhiyuan Chen, Bing Liu, Xiaokai Wei, Jidong Shao
  IEEE International Conference on Data Mining (ICDM'14), December 14-17, 2014 - Shenzhen, China.
  
  
• CatchSync: Catching Synchronized Behavior in Large Directed Graphs
  Meng Jiang, Peng Cui, Alex Beutel, Christos Faloutsos, Shiqiang Yang
  Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2014.
  
  
• Opinion Fraud Detection in Online Reviews using Network Effects  
  Leman Akoglu, Rishi Chandy, Christos Faloutsos
  Proceedings of The International AAAI Conference on Weblogs and Social Media (ICWSM-2013), July 8-10, 2013, Boston, USA
  
  
• CopyCatch: Stopping Group Attacks by Spotting Lockstep Behavior in Social Networks
  Alex Beutel, Wanhong Xu, Venkatesan Guruswami, Christopher Palow, Christos Faloutsos
  Proceedings of the 22nd International Conference on World Wide Web (WWW), 2013.
  
  
• What Yelp Fake Review Filter Might Be Doing 
  Arjun Mukherjee, Vivek Venkataraman, Bing Liu, and Natalie Glance
  Proceedings of The International AAAI Conference on Weblogs and Social Media (ICWSM-2013), July 8-10, 2013, Boston, USA.
  
  
• Exploiting Burstiness in Reviews for Review Spammer Detection
  Geli Fei, Arjun Mukherjee, Bing Liu, Meichun Hsu, Malu Castellanos, and Riddhiman Ghosh
  Proceedings of The International AAAI Conference on Weblogs and Social Media (ICWSM-2013), July 8-10, 2013, Boston, USA
  
  
• Review Spam Detection via Temporal Pattern Discovery
  Sihong Xie, Guan Wang, Shuyang Lin, Philip S Yu
  Proc. ACM KDD Conference, Beijing, China, Aug. 2012.
  
  
• Syntactic Stylometry for Deception Detection.
  Song Feng, Ritwik Banerjee and Yejin Choi. 
  Association for Computational Linguistics (ACL), 2012. 
  
  
• Finding Deceptive Opinion Spam by Any Stretch of the Imagination.
  Myle Ott, Yejin Choi, Claire Cardie, and Jeffrey Hancock. 
  Association for Computational Linguistics (ACL), 2011.

3.4. Workshops

• 2020 Workshop on Anomaly Detection in Finance
  co-organized in conjunction with SIGKDD 2020
• 2019 Workshop on Anomaly Detection in Finance
  co-organized in conjunction with SIGKDD 2019
• 2018 Workshop ODD v5.0 on de-constructing outlier detection methods
  co-organized in conjunction with SIGKDD 2018
• 2016 Workshop ODD 4.0 on outlier definition, detection and description on demand
  co-organized in conjunction with SIGKDD 2016
• 2015 Workshop ODDx3 on outlier definition, detection and description
  co-organized in conjunction with SIGKDD 2015
• 2014 Workshop ODD^2 on outlier detection and description under data diversity
  co-organized in conjunction with SIGKDD 2014

3.5. Tutorials

• Forecasting Big Time Series: Theory and Practice,
  Christos Faloutsos, Valentin Flunkert, Jan Gasthaus, Tim Januschowski and Yuyang (Bernie) Wang,
  TheWebConf (aka WWW), April 2020, Taipei, Taiwan.
  
  
• Fact checking: theory and practice,
  Xin Luna Dong, Christos Faloutsos, Xian Li, Subhabrata Mukherjee, and Prashant Shiralkar,
  KDD 2018, London UK, Aug. 19-23, 2018.


• Forecasting Big Time Series: Old and New
  Christos Faloutsos, Jan Gasthaus, Tim Januschowski and Yuyang Wang,
  VLDB 2018, Rio De Janeiro, Brazil, Aug. 27-31, 2018.


• Graph and Tensor Mining for Fun and Profit ,
  Xin Luna Dong, Christos Faloutsos, Andrey Kan, Subhabrata Mukherjee and Jun Ma.
  KDD 2018, London UK, Aug. 19-23, 2018.


• Data-Driven Approaches towards Malicious Behavior Modeling
  Meng Jiang, Srijan Kumar, VS Subrahmanian, and Christos Faloutsos.
  KDD 2017, Halifax, Nova Scotia, Canada, Aug. 13-17, 2017.

• Fraud detection through graph-based user behavior modeling.  
  Alex Beutel, Leman Akoglu, Christos Faloutsos.
  Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security Tutorial, Denver, Colorado, US, 2015
  
  

• Graph-Based User Behavior Modeling: From Prediction to Fraud Detection  
  Alex Beutel, Leman Akoglu, Christos Faloutsos.
  ACM SIGKDD Tutorial, Sydney, AU, August 2015 (3h-tutorial).
  
  

• Mining and Forecasting of Big Time-series data  
  Yasushi Sakurai,Yasuko Matsubara, Christos Faloutsos.
  ACM SIGMOD Tutorial, Melbourne, AU, May 2015 (3h-tutorial).
  
  

4. EDUCATION

The educational contributions of the project include the following dissertations:

• Dhivya Eswaran, Mining Anomalies using Static and Dynamic Graphs, April 2020.
  
• Junting Ye, Network Analysis and Modeling for News and Social Media, April 2019.
  
• Neil Shah, Anomaly Detection in Large Social Graphs. October, 2017.
• Shebuti Rayana, Ensemble and Multimodal Learning for Anomaly Mining: Algorithms and Applications. August, 2017.
• Huayi Li, Detecting Opinion Spam in Commercial Review Websites. August, 2016.
• Alex Beutel, User Behavior Modeling with Large-Scale Graph Analysis. May, 2016. (KDD dissertation award - runner-up)
• Danai Koutra, Exploring and Making Sense of Large Graphs. Dissertation, CMU, August 2015. (KDD dissertation award).

Point of Contact: Leman Akoglu, lakoglu AT cs.cmu.edu

---

Last updated: May 2020, by Leman Akoglu