Infoscience

Report

Survey of Preference Elicitation Methods

As people increasingly rely on interactive decision support systems to choose products and make decisions, building effective interfaces for these systems becomes more and more challenging due to the explosion of on-line information, the initial incomplete user preference and user's cognitive and emotional limitations of information processing. How to accurately elicit user's preference hence becomes the main concern of current decision support systems. This paper is a survey of the typical preference elicitation methods proposed by related research works, starting from the traditional utility function elicitation method and analytic hierarchy process method, to computer aided elicitation methods including example critiquing, needs-oriented, comparison matrix, CP-network, clustering & matching and collaborative filtering.

Related material

Contacts

EPFL authors