In this research some of the most significant capabilities of data mining techniques were leveraged through a multi-purpose framework for intelligent crime investigation. The framework exploited a systematic approach for using SOM and MLP neural networks for clustering and classifying crime data. Considerations and challenges of using hierarchical/partitional clustering techniques in crime data clustering were also discussed. We hope we can make crime spatio-temporal data involved in the analysis besides behavioral crime variables. We also intend to implement this framework as an integrated enterprise software.