The Research Of Intelligentize Business Based Telecom's Custom Data

Data mining is the process of revealing instructive but buried information and knowledge hidden in a large. incomplete, fuzzy, stochastic data. It always includes business understanding, data preparation, model building and results evaluation.After several years research. data mining algorithms is now mature and steady. The focus has turned to application domain. In telecommunication industry, when implementing data mining process analyst should fully understand business and process of the domain.This paper introduces basic concepts and technology of data mining, and discusses several standard data mining process model. Based on experience from practical mining work in telecom domain, we present the process of business understanding, data understanding, data preparation in data mining in the domain.First of all, comprehend the general business knowledge and process in the telecom domain. From the essential question in business, "how to gain profit", we practice a questioning process: divide a general question into several sub-questions, and continue divide sub-questions until acquire a collection of business questions; sum up analytical factors; strike up the mapping relationship between the factors and physical data such as table and field in database. In the later part of this thesis, a process of mining chum in teleco industri is implemented.