CPrefMiner: A Bayesian Miner of Conditional Preferences

Abstract

Customizing database queries by considering  user preferences is a  research topic that has been raisinga lot of interest within the database community in recent years.  Such preferences are used for sorting and selecting the best tuples, those which most fulfill the user wishes. A topic of interest within this context is the elicitation of preferences, consisting of methods to enable the user toinform his choice on pairs of objects belonging to a database. Depending on the size of the database, this task may requirea great effort from the user, and consequently  may discourage him/her to use the system. In this paper, we propose a first step towards the design and implementation of an automatic tool for inferring preferences from a given sample of user preferences. The method CPrefMiner we propose is based on the framework of Bayesian Networks and  aims at mining a special kind of preferences, the conditional preferences . The two main learning tasks accomplished by CPrefMiner are: (1) learning the graph underlying the conditional preference network; (2) learning the preference probability tables associated with each node of the graph. This paper focuses on the first task.