In the era of Internet, huge amounts of data are available to everybody, in every place and at any moment. As more information becomes available, it becomes increasingly difficult to search for relevant information: the main challenge is to support Web users in order to improve searching among extremely large Web repositories, such as online product catalogues or other generic information sources. Building systems for assisting users in finding relevant information is often complicated by the difficulty in articulating user interests in a form that can be used for searching. Machine learning methods offer a promising approach to solve this problem. Our research focuses on methods for learning user profiles which are predictively accurate and comprehensible. In this paper we present a comparison between an ILP and a probabilistic approach to learning models of users’ preferences. Experimental results highlight the usefulness and drawbacks of each one.