Abstract: A knowledge tree building system iteratively partitions a database of case records into a tree of conceptually meaningful clusters. Each cluster is automatically assigned a unique conceptual meaning in accordance with its unique pattern of typicality and exceptionality within the knowledge tree; no prior domain-dependent knowledge is required. The system fully utilizes all available quantitative and qualitative case record data. Knowledge trees built by the system are particularly well suited for artificial intelligence applications such as pattern classification and nonmonotonic reasoning.