Understanding the process of categorization is of great importance for building intelligent agents. Formulated categories help agents find information easier and understand the external world better. Instance-based categorization and prototype-based categorization have been two dominant approaches in the AI community. However, they share some drawbacks in common. First, they are crisp boundarybased hard categorizations (similar to classification). Second, they are not well-suited for dynamic category learning and formation. In this paper, we propose a hybrid soft categorization in the conceptual level that overcomes these drawbacks. The hybrid soft categorization merges the two popular hard categorizations and provides a robust fuzzy boundary-based soft categorization.