Decision support systems often require knowledge of users' preferences . However , preferences may vary among individual users or be dif ficult for users to articulate . This paper describes how user preferences can be acquired in the form of preference predicates by a learning apprentice system and proposes two new instance-based algorithms for preference predicate acquisition : 1 ARC and Compositional Instance -Based Learning (CIBL) . An empirical evaluation using simulated preference behavior indicated that the instance-based approaches are preferable to decision-tree induction and perceptrons as the learning component of a learning apprentice system , if representation of the relevant characteristics of problem-solving states , requires a large number of attributes , if attributes interact in a complex fashion , or if there are very few training instances . Conversely , decision-tree induction or perceptron learning is preferable if there are a small number of attributes and the attributes do not interact in a complex fashion unless there are very few training instances . When tested as the learning component of a learning apprentice system used by astronomers for scheduling astronomical observations , both CIBL and decision-tree induction rapidly achieved useful levels of accuracy in predicting the astronomers' preferences .