Acquiring Utility Functions from Holistic Comparisons

by Giuseppe Carenini

Combining data exploration and utility elicitation offers great potential to support users in the common and often challenging decision of selecting a preferred entity out of a possibly large set of available alternatives. In this talk, I describe some preliminary ideas on how an additive utility function for a user can be acquired by learning from holistic judgments that we assume could be naturally expressed by the user while exploring a set of alternatives to accomplish the selection decision task.

Back to the LCI Forum page