Modelling people's preferences
We assume a model for people's preferences which we can then use to extrapolate their true preferences from the data we receive from them. The model we will be using, the TM model, assumes that we can assign a normal probability distribution over some utility value for any given alternative. The applet below allows you to explore the model used to represent people's preferences.
The aggregation phase
The next applet illustrates the process of aggregating people's preferences. In particular, it highlights two special aggregations rules, Borda count and the lexmin function.