Several Studies About Business Data Processing

Data mining and text mining are very effective technologies and methods among business data processing applications, and they are also representatives of high technologies in business data processing; presentation / delivery of results of data processing can not be absent in a complete solution for business data processing; simultaneously, the above three aspects are also the most widely applied in Business Intelligence applications, and if these three aspects are successfully realized, that is to say, the basic frame of Business Intelligence applications and the main part of the solutions for Business Intelligence applications are constructed and perfected, too; and much exploring space still exists in the studies and applications of these three aspects currently.This dissertation deeply and completely studies and practises these three aspects, and based on digesting, absorbing and applying all kinds of (new) technologies it dissertates many methods for applying these three aspects, and brings forward a complete solution, and takes them as application paradigms to dissertate how to perfect applications at these three aspects.This dissertation points out shortages in current researches and applications about data mining, for example, the algorithms in the applications are relatively single; the understandings of the concepts, theories and procedures of data mining are still not profound enough, even some misunderstandings exist in this area; and the understandings, analysis and applications about data mining results are not enough. Aiming at these shortages, this dissertation discusses the theory and technology foundations and related concepts of data mining briefly, and analyzes and compares many data mining algorithms, dissertates the rules for data mining system selection, and supplies several golden rules for data mining applications.This dissertation deeply studies and analyzes some important techniques about preprocessing for text mining and techniques / principles about text classification, text clustering, automatic summarization for document(s) and other fields of text mining. While dissertating these techniques the back part of this dissertation echoes the frontal part, and the dissertation bands some of these techniques together while dissertating them, and makes the train of thoughts for text mining clear.This dissertation puts forward 5 important problems that desiderate resolution about presentations and deliveries for business information. Aiming at these problems, this dissertation discusses many techniques for presentations and deliveries about business information, a strategy about using information delivery techniques, and focuses on dissertating the techniques of Reprting Services and Notification Services of SQL Server 2005, these two kinds of techniques can support the application developments very well for presentations and deliveries about business information.Based on the above theoretic analysises about the related technologies, this dissertation discusses the understandings, analysises and applications about data mining results and evaluation skills for data mining models in detail. This dissertation discusses the applications in detail for realizing preprocessing about text mining, classification mining about English text, automatic classification about Chinese text and clustering mining about English text. This dissertation uses SQL Server 2005 as the development tool (especially implements Reporting Services), integrating Microsoft SharePoint Services, the techniques of the application and management of Windows domain and of Website developments, to develop a complete and feasible application project for representing and deliverying business information and other information, and dissertates the related key techniques in detail. Through practical operations and experimental verification, this project gains the scheduled ends, and solves several problems existing in this field; and it can completely be applied in the related projects of many organizations.The significances of this dissertation are: deepening the studies and applications about data mining, text mining and presentations and deliveries for business information; making up the former shortages in these aspects; and bringing forward different research hotspots for these aspects again.