In many application domains, there is a need for learning algorithms that generate accurate as well as comprehensible classifiers. In this paper, we present TRIPPER - a rule induction algorithm that extends RIPPER, a widely used rule-learning algorithm. TRIPPER exploits background knowledge in the form of taxonomies over values of features used to describe data. We compare the performance of TRIPPER with that of RIPPER on a text classification problem (using the Reuters 21578 dataset). Experiments were performed using WordNet (a human-generated taxonomy), as well as a taxonomy generated by WTL (Word Taxonomy Learning) algorithm. Our experiments show that the rules generated by TRIPPER are generally more accurate and more concise (and hence more comprehensible) than those generated by RIPPER.