Document Type



In this paper, we have developed a data warehouse and an OLAP( on-line analytical processing ) based framework that has been used for customer profiling and comparison. The system architecture for OLAP and data warehouse based calling behavior profiling and multilevel multidimensional pattern analysis is introduced. First, the method how customer profiles can be represented as data cubes is described, then, the architecture of a profiling engine is presented, finally, the process of using the engine to compute profiles and calling patterns is discussed and an application case of China telecom is studied.