Selecting news to suit a group of criteria: An exploration

In (Masthoff, 2004), we have investigated techniques for combining individual user models to make recommendations to a group. In Masthoff (2003), we have shown how these techniques might also be applicable when adapting to an individual: for solving the cold start problem and for combining ratings on multiple criteria. In this paper, we look at combining multiple criteria in more detail. We present an exploratory experiment into how people combine criteria. A main issue is potential inequality of the criteria’s importance. We show how the group modelling strategies can be adapted to deal with this inequality.