The Research Of Date Warehouse Of Common Data Define

This paper discusses the methodologys, processes and related templates and products that are used to create a set of common data standard before building a data warehouse system, which is based on CCB's "common data standard project"completed in the prophase of creating a data warehouse system. The common data standard is to make a uniform data standard for the data shared by CCB's business lines. It consists of the business data standard and the technology data standard. The common business data standard is business characteristics of the common data, such as the definition, classification and rules of business. Its objective is to make the business department have a unanimous understanding and application principle. The common technology data standard is technology characteristics of the common data, such as the type, format and configuration of data. It is to find out the data representation in the system for the business requirement. Its objective is to make the technology department have the same basic standard to data definition during the system development. The whole methodology to do this work includes four major stages, which are the data requirement definition, the data standard list production, the data standard establishing and the data standard examining & approving. The data requirement definition The data requirements in the data standard stem from recertified multi-data model, KPI and other business functions and existing business demand. A wide range of data models include CCB's core logic data model, client model, product model, internal structure model, internal financial account structure model. Models, KPI and other business functions and existing business functions mentioned above can provide part of the data requirement. Otherwise, the data requirements have used the optimized industry practice and the business data model adopted at large in the industry field for conference. The actual methodology to define the requirements is as follows. ? The produce of data standard list First, data items in the model will be extracted out with CCB's core logic data model which is identified. Second, business data items which need to be defined by business is selected out with the prepared methodology to select business data items. At last, the selected business data items are classified by business, and the data item list under data standard for the common data standard and the common code definition is created. The sample of this working methodology is as follows. In this paper, business rules are discussed in detail with business logic models for the supervising principle in the process of making the common data standard. ? The data standard establishing The actual methodologys to define the data standard consist of two parts, which are the methodology to define the common data standard and the methodology to define the common code. Description of the working process: 1. According to the natural classification methodology of data items (to be explained below), classify the data items in the common data standard list. 2. Based on the results in the natural classification, according to the methodology to define the business data standard (to be explained below), define the business characteristics for these data item classification. 3. Based on the definition of the business characteristics, specify the requirements of data item technical characters caused by business requirement, andcreate the definition of the technology standard. 4. Organize the review within the team, professional, and CCB's related department. n The methodology to define the common code The working object of the common code is primary to define the actual content of the common code. The common code definition includes two parts. 1. The business meanings and background of code itself. 2. The range of values to set the code content effectively: based on analyzing and classifying the actual business information, specify the illustration of the actual code, coding principle and code meanings. n The supervising principle to define the common code The tiptop level of the supervising principle to define the common code is showed as follows, the principles align in the order of importance. It provides the basic and guide for the detail principles to define the common code. 1. Aim at meeting the international standard, and apply the international standard in sufficiency/maximum. 2. Based on the international standard and consider the specific condition and requirement of national area, business field and industry cycle, redefine, extend and alter the adopted standard. 3. As to identify the consult system and the definition base of the common code, it will be considered clearly that includes the representation of the consult system and the comprehension and applicability of its code, and the extension ability to design for the future demand. 4. The definition of client oriented common code may use the related industry experience and CCB's future business demand as the major input. 5. If there are many coding ways in the national or international standard, it is like to adopt the data type coding. If there is no data type coding, it is better to adopta certain coding way that considers the difficulty degree of coding and relates to CCB's practice. n Working process to define the common code the content of commom code work can divide into 4 stages: 1Common code source classfication According to public code content, using public code classfication template divide into the code source that defined according to content existing unified standard code ( international, national standard code ,CCB's unified code ), source systematic code and other codes 2 Collect and confirm that is international , national standard code and CCB's unified code Collecting international, national standard code and CCB's unified code content by different channel By contrast and analysis related code table content from source system, to adjust and confirm the content of code by discuss with business department . ( instead of apply mechanically international, national standard code totally, depend on whether suit to CCB's busyniss requirement ) 3 Collect and analyse source systematic code content , and unify definition The source code content is that current content and the code table in CCB's every system Major work is analys source system data dictionary, arrange the code table name needed, and extract the content of related table by source system manager. Analyse the content of source system code table that had gotten , exact its content and the code table concerning common code For many source systematic code tables that have identical meaning,, define a unified code content and meaning by discuss with related business department. 4 Collect and define other code contents Other codes is never defined by international , country or CCB. Major work is to collect through different channel for code content ( professionbest practice , accenture company's knowledge storehouse and so on) For the information that collected , analyse and discuss , define code content eventually At last, it's to review the created common data standard within the team, by the professionals and through the business department. ? The data standard examining and approving After making the common data standard and the common code, it must be validated through the process as follows: 1 Review the created common data standard within the team It's means to form unified opinion about data standard content within the team 2 External expert review the common data standard Accenture Business expert basis knowledge and the experience of themself in business field the best practice that combines trade circle for data standard offer feedback opinion 3 Internal expert review the common data standard CCB's business department basis the demand for service of each business field and the specific business of CCB's process for data standard offer feedback opinion. 4 Data standard is tested and is popularized Project group will carry out exchange with consummately data standard and each related project group through PMO, since the angle number of application offers feedback opinion according to standard To meet requirements of the business management, the business management information which CCB needs is classified to 12 big species (or named "information species"). The information species include "group", "vital statistics", "address andgeography position", "production", "account", "application", "marketing", "channel", "event", "model", "risk", "general ledger". It is described in detail in this project for the content and business rules of each information species, and the relationship among every information species. Above all, this project, as the prophase preparation of the data warehouse, uses the experience in the business field for conference, and creates CCB's common data standard through the perfect working methodology and the precise working process. It can be considered as an example to ensure the data warehouse with the data consistency and unify the business system.