Personalisation & Social Network Analysis Group
Organisations and individuals produce information, which is largely accessible at the Web, at an unprecedented rate. Yet human capacity to consume and comprehend information has not changed, which leads to the phenomenon of information overload. The field of personalisation aims to address the said problem, by filtering the information that is available according to the needs of individual users. In doing so, the users are represented by profiles (or models) which are matched against profiles of information. Building user profiles can be performed by the users themselves which is feasible to a certain extent only, but inferring the user profiles by probing the users’ behaviour is both more convenient and more interesting. The profiles represent the learnt preferences of users and can be based on their past behaviour with respect to web sites, or their similarity to other users.
A rather recent phenomenon, is the enormous growth (as attested by user participation and content creation) of on-line communities of various forms. Each user is no longer isolated but dependent on the links formed to other users, thus forming a social network. Therefore, the issue of information personalisation is rendered more interested and precise by taking into account not only the behaviour of a user but also his connections.
The information personalisation and social network analysis (PerSoN) group of NCSR "Demokritos" works towards user modelling, and information personalisation in recommender systems that suggest items of interest to the users that range from books, and fashion apparels up to university courses. The PerSoN group has produced and is maintaining a personalisation server in the form of open source software that aids the maintenance of atomic profiles, but also the discovery of collective profiles. Furthermore the PerSoN group has an interest in community detection methods as well as in trustworthiness discovery in social networks.
The PerSoNA group is part of the Software & Knowledge Engineering Lab of the Institute of Informatics and Telecommunications in the National Centre for Scientific Research "Demokritos".