OnSt ’17: The 2nd International Workshop on Online Safety, Trust and Fraud Prevention
Collocated with ECIR 2017, the 39th European Conference on Information Retrieval
Almost every aspect of our lives is influenced by the web, such as in entertainment, education, health, commerce, the government, social interaction, and many more. The profound impact these changes have on our lives requires to rethink the way we make decisions in these areas.
Besides questions related to cost and benefit, there are important issues raised by users regarding trust and safety. Can I trust this service with my data? Is it safe to use? In many cases, trust-related issues become a dealbreaker for the adoption of online services. We commonly find cases where people avoid online banking or buying products online due to the fear of becoming prey of fraudulent activity. Thus, providing a trustworthy environment for users is of utmost importance.
When online fraud are committed, fraudsters take advantage of gaps allowing them to unjustifiably enrich themselves. However, when fighting fraud, companies face a dilemma, given that no system is perfect: in e-commerce, on the one hand, fraud and its related losses should be reduced; on the other hand, users neither want to be accused of fraud nor treated like criminals. In other areas, such as health, the problems associated with data abuse and security leaks could even result in more severe damage than purely financial matters.
Yet, companies’ practical implementations of fraud investigation processes rarely meet scientific standards. Fraud prevention companies offer diverse products on the market, but neither their effectiveness nor their efficiency has been verified scientifically until now. Dealing with these issues, the first step should be a clear definition of what constitutes fraud and how it can be measured. Afterwards, models can be built and tested according to scientific standards of evaluation. On this basis, a careful risk analysis can be conducted in order to weigh pros and cons of pursuing individual suspicious cases.
Companies spend millions to protect themselves from fraudulent activities. One of the most interesting aspects is the fact that fighting fraud is a social interaction that needs constant supervision and improvement: a never ending race between criminals and investigators, in which both actors adjust for the actions of the other.
This workshop aims at bringing together researchers from a wide range of disciplines (mathematics, computer science, economy, philosophy, social science) to (i) understand the cases and motivations of fraudulent activities in online environments, (ii) find solutions to detect and analyze fraud, and (iii) derive means to prevent it.
The OnST workshop focuses on online fraud detection and prevention, but submissions that tackle these challenges in other environments are welcomed as well. Relevant research areas to the workshop are, e.g., “Spam detection”, “Trust, authority, reputation, ranking”, and “Time series and forecasting” with the goal of anomaly detection. We invite the submission of on-going and mature research work with a particular focus on the following topics:
- Online Safety and Trust
- User Modeling: personalization of fraudulent and malicious users;
- Account takeover;
- Human interactions;
- IT-forensics investigating a wide variety of crime, including child pornography, fraud, espionage, cyber-stalking, etc.
- Fraud Prevention
- Feature engineering for online detection;
- Supervised machine learning techniques: fraud rule engines, time series, spatial-based, graph-based, spatio-temporal approaches;
- Unsupervised machine learning techniques: Outlier and anomaly detection;
- In Crowdfunding;
- In Big Streaming Data;
- Distributed systems;
- Effective and efficient systems.
The main research questions that the workshop would like to answer are:
We invite the submission of original work in these and related areas. Each submission to the workshop will be peer-reviewed by at least two expert reviewers.
Workshop Rationale and Significance and Relevance to ECIR
The proposed workshop has a strong relation to most topics of the main program of ECIR, such as “Spam detection”, “Trust, authority, reputation, ranking”, and “Time series and forecasting” with the goal of anomaly detection. As such, the workshop will be widely accessible to the ECIR community. At the same time, though, this workshop will approach the above ECIR topics from the unique and emerging viewpoint of detecting and preventing malicious activities from a scientific point of view, considering their psychological and , economical impact as well as their risk. This viewpoint and the workshop’s focus clearly differentiate the workshop from ECIR’s main program and make it an appealing addition to it.
The workshop will be held on the 9th of April, starting at 9.05 am, at Robert Gordon University, in the Sir Ian Wood Building, Garthdee Road, Aberdeen, AB10 7GJ.:
09:05 – 09:10 Welcome and Introduction
09:10 – 10:00 An Economic Perspective on Fraud Prevention in the E-Commerce Mailorder Business – Tobias Knuth
10:00 – 10:30 Break
10:30 – 11:20 Temporal Anomaly Detection in the Retailer Sector – Steffen Brauer
11:20 – 12:30 Round Tables about Online Safety, Trust and Fraud Prevention
New deadline for submissions: 17 February 2017
Notification of acceptance: 10 March 2017
Camera-ready: 17 March 2017
Workshop date: 9 April 2017
Dr. Marco Fisichella is the head of the data science team at Risk Ident (Otto group, Germany) where they devise algorithms in order to detect the online frauds. Before he joined Risk Ident, he was postdoctoral researcher at the L3S Research Center in Hannover, Germany. Until beginning of 2015, he was also lecturer of the Artificial Intelligence course for the Master in Computer Science at the Leibniz University of Hannover. His research interests include data mining, information retrieval, generative model, event detection, clustering methods based on statistical approaches, near duplicate detection. He has worked as project manager in several EUfunded projects including (1) DuraArk Preservation of architectural building data, and (2) OpenScout accelerating the use, improvement and distribution of open content in the field of management education and training. He actively participated as proposal consultant and advisory board member in the following accepted proposal: (1) ALEXANDRIA an ERC Advanced Grant Project on Foundations for Temporal Retrieval, Exploration and Analytics in Web Archives; (2) Zivile Sicherheit a BMBF German funding program for tracking the Russian flu in U.S. and German medical and popular reports, occurred between 1889 and 1893. He has strong publication records in toptier conferences, such as CIKM, SPIRE, ECIR and WISE and active professional memberships, i.e., invited reviewer and PC member (e.g., ICDM, WWW, and CIKM) and journal reviewer (e.g., Data & Knowledge Engineering Journal Elsevier on the track area about Reasoning Approaches). Finally, he received the best paper award at ECTEL 2011 for his publication on “Unsupervised Autotagging for Learning Object Enrichment”
Prof. Dr. Nattiya Kanhabua is an assistant professor at the Department of Computer Science, Aalborg University, Denmark. Her research interests are information retrieval, data mining, machine learning, and spatial and temporal analytics. She did her PhD at the Department of Computer and Information Science, Norwegian University of Science and Technology (NTNU). She was a postdoctoral researcher at the L3S Research Center Hannover, Germany. At L3S, she worked in several research projects, e.g., 1) EU Project ForgetIT: Concise Preservation by Combining Managed Forgetting and Contextualized Remembering, 2) ALEXANDRIA, an ERC Advanced Grant Project on Foundations for Temporal Retrieval, Exploration and Analytics in Web Archives, and 3) Medical Ecosystem: Personalized Eventbased Surveillance. She has published her research work in toptier conferences, e.g., SIGIR, WSDM, CIKM, JCDL and ECIR.
Sven Kurras is a senior data scientist at Risk Ident (Otto group, Germany). He focuses on detecting fraudulent behavior from connectivity information by developing scalable techniques for statistical graphbased inference. Beside his work at Risk Ident he finishes his doctoral thesis to the end of 2016 within the DFG Research Unit 1735 “Structural Inference in Statistics: Adaptation and Efficiency”. During the preceding four years, he worked fulltime as a doctoral researcher in the field of unsupervised machine learning at the machine learning working group of Ulrike von Luxburg at the University of Hamburg. He specialized on theoretical foundations of multiscale clustering algorithms on random graphs. His results are published at international topconferences like ICML and AISTATS. Complementary to his theoretical background, he also brings in 15 years of experience as a Java software architect and programmer, recently shifting to Scala and big data architectures such as the SMACK stack.
- Prof. Dr. Ismail Sengor Altingovde, Dept. of Computer Engineering Middle East Technical University (METU), Turkey
- Steffen Brauer, Risk.Ident GmbH Otto Group, Germany
- Andrea Ceroni, L3S Research Center Leibniz University Hannover, Germany
- Marco Diciolla, Palantir Technologies, U.K.
- Dr. Marco Fisichella, Risk.Ident GmbH Otto Group, Germany
- Prof. Dr. Felix Freiling, Friedrich-Alexander-Universität Erlangen-Nürnberg, Germany
- Sergio Govoni, ProQuest, U.S.A.
- Prof. Dr. Rüdiger Grimm, Fellow of the German Informatics Society GI e.V., Germany
- Prof. Dr. Nattiya Kanhabua, Aalborg University, Denmark
- Sven Kurras, Risk.Ident GmbH Otto Group, Germany
- David Losada, University of Santiago de Compostela, Spain
- Dr. Ida Mele, Faculty of Informatics - Università della Svizzera italiana (USI), Switzerland
- Dr. Katja Niemann, Fraunhofer Institute for Applied Information Technology (FIT), Germany
- Simon Schenk, Risk.Ident GmbH Otto Group, Germany
The entire workshop will gain and leverage from the hard grounded experience of (i) Prof. Dr. Rüdiger Grimm, who was also head of the research group “Security for Virtual Goods” of the Fraunhofer Institute for Digital Media Technology; and (ii) Prof. Dr. Felix Freiling, former advisor of the German constitutional court (Bundesverfassungsgericht) in cases related to data protection and police laws in Germany.
Prof. Dr. Rüdiger Grimm was professor for IT Risk Management at the University Koblenz-Landau since 2005-2015, and he is continuing research and teaching duties in his University after his retirement in October 2015. During that time, 2002-2005 he was also head of the research group “Security for Virtual Goods” of the Fraunhofer Institute for Digital Media Technology (IDMT) in Ilmenau. 2011-2014 he was elected Dean of the Faculty of Informatics in Koblenz. Since 2010 he is Fellow of the German Informatics Society GI e.V.
Prof. Dr. Felix Freiling is a full professor of computer science at Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) in Erlangen, Germany. Before joining FAU he held professor positions at RWTH Aachen University and University of Mannheim. His research interests are in digital forensics and offensive computer security. He is member of the Steering Committee of the International Conference on IT Security Incident Management & IT Forensics (IMF) and was the chair of the TPC of the Digital Forensics Research Conference Europe (DFRWS EU) 2015. He was an advisor of the German constitutional court (Bundesverfassungsgericht) in cases related to data protection and police laws in Germany.