SoMeRis 2016

2nd International Workshop on Data Science for Social Media and Risk

December 12, 2016    Barcelona, Spain

supported by the AXA-Polytechnique DASCIS chair


Organizers


Check photos of the SoMeRiS 2016 workshop. Check also the 1st SoMeRis workshop, held in conjunction with ASONAM'15. Discover other ICDM 2016 workshops here.


Find Out More

Introduction

The topic of risk assessment deals with methods, quantitative or qualitative, to estimate the risk that is related to a specific task or action. As life becomes increasingly digital, the available social media data - including interactions, time, location and behavior - is constantly increasing, giving rise to quantitative methods from the broad field of data science towards "better" risk assessment. The main goal of the SoMeRis workshop is to act as a venue for research methods and tools on the emerging field of data science and social media analytics for risk assessment and management.

Call for Papers

The aim of this workshop is to address state-of-the-art techniques in machine learning, graph and text mining to leverage information available in social media to enrich and improve predictive algorithms for risk assessment and management, such as the insurance sector or the online industry. A (non restrictive) list of topics follows:

Risk management

  • Social media for risk assessment in online industry
  • Social media for risk in the insurance sector
  • The nature of risk in the insurance sector
  • Fusion of data sources for risk assessment and management

Social media

  • Graph/Text Mining for social media
  • Urban social media analysis and risk
  • Dynamics of social networks
  • Anomaly detection in social media

Privacy

  • Privacy in Social media
  • Fraud detection and prevention
  • Sharing economics, social networks and peer-to-peer insurance
  • Benchmarks for social media mining and risk

The workshop will host two keynote speakers (one from a major insurance company (AXA) and one from academia), several academic and industrial talks as well as a discussion panel on the importance of social media in risk management in the insurance sector. The objective is to create an interface between academics and practitioners, to exchange and create synergies about challenging problems at the crossing of machine learning on social media and industrial applications that involve risk assessment and management. This topic is particularly relevant and unexplored. The exploitation of such sources of information for risk-based businesses is a relatively recent topic. It requires new developments both on the methodological and technological aspects: from new statistical models that describe accurately the dynamics of social media and information diffusion, to the use of APIs and relevant databases to crawl, structure and exploit data from social media. Needless to stress that the dominant structures in this case are the networks enriched with the textual content of user interactions and histories.

The expected added value of this workshop will be:
  • The initiation of discussion between academics and practitioners, towards the identification of common grounds and understanding of the risk management on the context of social media for online services.
  • The creation of an event bridging social media with the insurance industry – a case were risk management is very important. In this context it is expected that this synergy will develop new approaches for risk management put in the relevant context.

Key Dates

August 12, 2016 August 26, 2016: Due date for full workshop papers

September 13, 2016 September 15, 2016: Notification of workshop papers acceptance to authors

September 20, 2016: Camera-ready deadline for accepted papers

December 12, 2016: Workshop date

ACCEPTED PAPERS

Graph-based Term Weighting Scheme for Topic Modeling
Giannis Bekoulis (LIX Ecole Polytechnique, France)
François Rousseau (LIX Ecole Polytechnique, France)

Investors Attention and the Effects on Stock market: An Empirical Study Based on Stock forum
Wen Long (Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, China)
Lijing Guan (Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, China)
Lingxiao Cui (Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences, China)

Identifying Warning Behaviors of Violent Lone Offenders in Written Communication
Lisa Kaati (Uppsala University, Sweden)
Amendra Shrestha (Uppsala University, Sweden)
Rachel Doveikis (Evolve 24, USA)
Seth Howell (Evolve 24, USA)
Tony Sardella (Evolve 24, USA)

Microblog Sentiment Topic Model
Aman Ahuja (BITS Pilani - K.K. Birla Goa Campus, India)
Wei Wei (Carnegie Mellon University, USA)
Kathleen Carley (Carnegie Mellon University, USA)

Learning from User Workflows for the Characterization and Prediction of Software Crashes
Chloé Adam (Laboratory MICS - CentraleSupélec, France)
Antoine Aliotti (GE Healthcare, France)
Paul-Henry Cournede (Laboratory MICS - CentraleSupélec, France)

Time-Based Ensembles for Prediction of Rare Events in News Streams
Nuno Moniz (LIAAD / INESC Tec, Portugal)
Luís Torgo (LIAAD / INESC Tec, Portugal)
Magdalini Eirinaki (Department of Computer Engineering - San Jose State University, United States)

Fraud Detection in Voice-based Identity Authentication Applications and Services
Saeid Safavi (University of Hertfordshire, United Kingdom)
Hock Gan (University of Hertfordshire, United Kingdom)
Iosif Mporas (University of Hertfordshire, United Kingdom)
Reza Sotudeh (University of Hertfordshire, United Kingdom)

Novel Probability Based Ensemble Method for Improving Automated Tweet Classification
Renhao Cui (The Ohio State University, United States)
Gagan Agrawal (The Ohio State University, United States)
Vinh Khuc (Astute Solutions, United States)
Rajiv Ramnath (The Ohio State University, United States)

Keynote


Jie Tang

Jie Tang

Associate Professor
Tsinghua University

Big Network Analysis—Algorithms, and Applications

Abstract: Online social networks connect our physical daily life and the virtual Web space. The user generated data is becoming big, heterogeneous, and highly connected. In this talk, I will first present our recently developed methodologies and algorithms for connecting multiple heterogeneous networks (COSNET) and top-k similarity search (Panther). Both algorithms have been deployed to an online academic search and mining system AMiner, which has collected a large scholar dataset, with more than 130,000,000 researcher profiles and 100,000,000 papers from multiple publication databases. With COSNET, we connect AMiner with several professional social networks, such as LinkedIn and VideoLectures, which significantly enriches the scholar metadata. Panther is used to find similar authors in AMiner and can return top-k similar vertices 300× faster than the state-of-the-art methods.

Bio: Jie Tang is a Tenured associate professor with the Department of Computer Science and Technology at Tsinghua University, and was also visiting scholar at Cornell University, Hong Kong University of Science and Technology, and Southampton University. His interests include social network analysis, data mining, and machine learning. He has published more than 200 journal/conference papers and holds 20 patents. His papers have been cited by more than 7,000 times (Google Scholar). He served as PC Co-Chair of CIKM’16, WSDM’15, ASONAM’15, SocInfo’12, KDD-CUP/Poster/Workshop/Local/Publication Co-Chair of KDD’11-15, and Associate Editor-in-Chief of ACM TKDD, Editors of IEEE TKDE/TBD and ACM TIST. He leads the project AMiner.org for academic social network analysis and mining, which has attracted more than 8 million independent IP accesses from 220 countries/regions in the world. He was honored with the UK Royal Society-Newton Advanced Fellowship Award, CCF Young Scientist Award, and NSFC Excellent Young Scholar.

You can find the slides: here.

Organization


Programme Committee

Jordi Casas Roma, Open University of Catalonia, Spain

Magdalini Eirinaki, San Jose State University, USA

Christos Giatsidis, École Polytechnique, France

Bruno Gonçalves, New York University, USA

Argyris Kalogeratos, ENS Cachan, France

Theodoros Lappas, Stevens Institute of Technology, USA

Clement Levallois, EMLYON Business School, France

Michael Mathioudakis, Aalto University, Finland

Kjetil Nørvåg, Norwegian University of Science and Technology, Norway

Evangelos Papalexakis, Carnegie Mellon University, USA

Jesse Read, Télécom ParisTech, France

Ryan Rossi, Palo Alto Research Center (PARC), USA

Ansaf Salleb-Aouissi, Columbia University, USA

Nikolaos Tziortziotis, École Polytechnique, France

Iraklis Varlamis, Harokopio University of Athens, Greece

Rui Yan, Baidu, China

Program

09:00 - 09:05
Opening of the Workshop

09:05 - 09:50
Keynote Speaker: Jie Tang, Tsinghua University

9:50 - 10:10
Time-Based Ensembles for Prediction of Rare Events in News Streams
Nuno Moniz, Luís Torgo, Magdalini Eirinaki

10:10 - 10:30
Fraud Detection in Voice-based Identity Authentication Applications and Services
Saeid Safavi, Hock Gan, Iosif Mporas, and Reza Sotudeh

10:30 - 11:00
Coffee - break

11:00 - 11:20
Identifying Warning Behaviors of Violent Lone Offenders in Written Communication
Lisa Kaati, Amendra Shrestha, Rachel Doveikis, Seth Howell, and Tony Sardella

11:20 - 11:40
Learning from User Workflows for the Characterization and Prediction of Software Crashes
Chloé Adam, Antoine Aliotti, and Paul-Henry Cournede

11:40 - 12:00
Investors Attention and the Effects on Stock market: An Empirical Study Based on Stock forum
Wen Long, Lijing Guan, and Lingxiao Cui

12:00 - 12:20
Microblog Sentiment Topic Model
Aman Ahuja, Wei Wei, and Kathleen Carley

12:20 - 12:40
Graph-based Term Weighting Scheme for Topic Modeling
Giannis Bekoulis and Francois Rousseau

12:40 - 13:00
Ensemble of Heterogeneous Classifiers for Improving Automated Tweet Classification
Renhao Cui, Gagan Agrawal, Vinh Khuc, and Rajiv Ramnath

13:00
Closing Remarks

Contact Us