A novel approach to news aggregation is proposed. Rather than ranking or summarisation of cluster topics, we propose that articles are grouped by topic similarity and then clustered within topic groups in order to identify archetypal articles that represent the various perspectives upon a topic. An example application is examined and a preliminary user study is discussed. Future applications and evaluation of validity are outlined.
|Title of host publication
|Proceeding BCS-HCI '12 Proceedings of the 26th Annual BCS Interaction Specialist Group Conference on People and Computers
|British Computer Society
|Published - 2012