摘要：In this paper, we propose a complex learning system for data analysis that is based on behavior factor. This system combines features of both order based systems and order based data analysis framework. This system uses an appropriate behavior factor acquisition process, which is the weakest task in the current order and behavior factor based systems. As several research works have shown that data analysis is rather driven by domain behavior factor than by data, we have identified and analyzed the characters that distinguish behavior factor and orders from data for better determining the most components of the proposed system. This paper describes an approach that advocates the use of behavior factor order-based systems techniques for data analysis systems. In particular, we aim to make domain behavior factor explicit and separate from the system where will be used for data analysis. For representing domain behavior factor, we investigated a uniform and unified order representation form based on the creation of the environment component that stored information especially for the management of order and its quality.
