This paper presents an approach for avoiding resource misallocations in Business Processes (BPs). We consider a resource misallocation as an improper combination of resources in a BP and we introduce a decision tree-based approach to minimize these misallocations likelihoods. The approach uses a BP's event log that tracks past misallocations along with their undesirable impacts in terms of time, cost, and quality on BP instances. These impacts are identified by mining resource allocation rules, which provide a concise summary of a BP's event log and thus are also used to generate decision trees. A system demonstrating the feasibility and efficiency of the approach through experiments is presented in this paper as well.