Special Session: Computational Issues in Voting Advice Applications

Last updated on  
  • May 28th, 2014, by
  • admin

Aims and scope

Voting Advice Applications (VAAs) are web applications, which are created to match voters' policy views with parties' or candidates' political positions and recommending a candidate or a party to a voter. Since smooth functioning of the democratic government depends on citizens' participation in public affairs, VAAs are considered as an important tool to mobilize participation in elections, taking advantage of their online social network nature. Although VAA research has been largely taken up by political scientists recently gained some attention by researchers working on the fields of artificial intelligence, information retrieval, data mining and social media. A tangible result of VAAs is the datasets that are created and which usually contain dozens of research variables (demographics, vote prediction, self placement in political maps, social network formation, etc.) each one requiring exhaustive investigation. VAA datasets are extensively used by political scientists but their utilization focuses mainly on strictly political science issues. Some, rare, exceptions simply confirm the above trend. VAA data, although rich in information and huge in volume, remain almost totally unexploited by computer scientists. Recently some activity has been reported towards this direction by the work of numerous researchers, however, there is a long way to go to establish it as research field in informatics.

This special session aims to bring together researchers from different fields working on computational issues of VAAs including VAA design issues (database, coding, interface), dimensionality reduction and visualization, VAA data analysis, VAA social network analysis, intelligent vote recommendation, etc. Furthermore, the Special Session organizer is a member of the steering committee of EUVox initiative who run a VAA across all European Union states for the European parliament elections in May 2014. Through EUVox an impressive VAA dataset containing the answers of one million users in 30 policy and 40 supplementary questions were recorded. This dataset is a very rich pool of information and can be utilized by researchers working in several fields. In the proposed special session we expect some initial research work on this dataset to be presented to further mobilize research interest in the area.


Topics of interest include, but are not limited to the following:

  • VAA designing issues (database design, party coding, VAA interface)
  • HCI issues in VAAs
  • Dimensionality reduction and visualization
  • VAA data analysis (data cleaning, outlier detection, vote prediction, missing value estimation)
  • VAA social network analysis
  • Measuring party coding effectiveness
  • VAA matching algorithms
  • Social Voting Recommendation
  • Data mining for party coding
  • Intelligent applications of VAA data


Dr. Nicolas Tsapatsoulis, Associate Professor,
Dept. of Communication and Internet Studies,
Cyprus University of Technology,
31 Archbishop Kyprianos Str., CY-3036, Limassol, Cyprus,
Tel.:+357 25002614, Fax: 25002664
Email: nicolas.tsapatsoulis@cut.ac.cy,
URL: http://www.cs.ucy.ac.cy/~nicolast/

Special Session Committee

  • Dinas E., University of Nottingham, UK
  • Gemenis K., University of Twente, Netherlands
  • Katakis I., National & Kapodistrian University of Athens, Greece
  • Mendez F., University of Zurich, Switzerland
  • Theodosiou Z., Cyprus University of Technology, Cyprus
  • Wheatley J., Centre for Research on Direct Democracy in Aarau, Switzerland