Identification of Unreliable Accounts in Social Networks - current
approaches and real-world scenarios
PRESENTER:
Nuno Guimaraes
University of Porto, Portugal
ABSTRACT:
The rise of social networks revolutionized the way people create, share,
and consume content. Consequently, these have become a platform where the
quantity of information is higher, travels quickly, and reaches a broader
audience compared with more traditional mediums (such as television and
radio). Social networks' popularity and lack of moderation allowed
unreliable information to proliferate and caused a significant impact in
the political, economic, and social domains.
This talk will be two-folded. In the first part, we attempt to summarize
the principal findings on the topic, highlighting the main research
efforts from the past few years to analyze and automatically identify
unreliable content and accounts. In the second part, we will focus on some
limitations of the current approaches in identifying unreliable accounts.
In addition, we will present some solutions (knowledge-based and
classification-based) that try to complement the current literature with a
more pragmatic approach capable of coping with real-world situations.