What Is a Data Mart?

Data Mart, also called data market, is to meet the needs of specific departments or users, and store it in a multi-dimensional manner, including defining dimensions, indicators that need to be calculated, levels of dimensions, etc. to generate decision-oriented Data cube for analysis needs.

1. Data marts are characterized by their small size.

The structure of data in a data mart is often described as a star or snowflake structure. A star structure contains two basic parts-a fact table and various supporting dimension tables.

In the well-known foreign Garnter report on data mart products, the agile business intelligence products in the first quadrant are QlikView, Tableau, and SpotView, all of which are full-memory computing data mart products. They formed a big data giant for traditional business intelligence products Challenged. Domestic BI products started late. Well-known agile business intelligence products include PowerBI, Yonghong Technology's Z-Suite,

When enterprises plan data warehouse projects, they often encounter many data warehouse software vendors. In addition to selling related software tools, various vendors also instill many concepts into the enterprise. among them,

In terms of data structure, a data warehouse is a collection of subject-oriented, integrated data. The data mart is usually defined as a star or snowflake data structure. The data mart is generally composed of a fact table and several dimension tables.

A data mart is a collection of data targeted at a specific set of subject areas, departments, or special user needs. This data needs to be optimized for quick user access and report display. The optimization methods include lightweight summary of the data and creation of an index based on the data structure. The data mart's target analysis process includes splitting the needs of the data mart, organizing according to different business rules, organizing entities related to business topics into topic domains, and performing dimensional analysis on various indicators to form data Bazaar target statement. The content includes detailed business topics, business topic fields, and various indicators and their analysis dimensions.

The cost of establishing data warehouses and data marts of different specifications is assessed by foreign consulting agencies and can be used for reference to a certain extent. but

The fast-growing, competitive business world has endless demand for timely and accurate information. Some IT experts believe that the inevitable result is the creation of data marts. Other experts question the work and costs of users and customers. After all, ca nt we just inherit

Through successful trials in cities such as Jilin City, China Mobile has decided to use data marts as one of the focus of mobile prefecture-level companies in 2006. This also means that the BI application based on the data warehouse in the telecommunications industry has entered a stage of deeper mining, and the results produced will directly serve the production and sales of the front line ^[5]

Data marts: the first step towards deeper mining

The telecommunications industry is no stranger to data warehouses. In order to achieve a transformation from product-oriented to customer-oriented, telecommunications companies have established customer-centric data warehouses in the hope that they will formulate strategies based on customer needs, expectations, and preferences to enhance corporate competitiveness. To put it simply, the data warehouse is to ensure the efficiency of data query and analysis. All data is stored in categories according to the theme. When needed, data can be extracted according to the theme and further analyzed and processed.

A data mart, which can be called a "small data warehouse", is a specialized collection of data used to analyze related specialized business problems or functional goals. It is built under a data warehouse with a unified data storage model. Business people at all levels replicate, process, and process data in accordance with the specific needs of each department, and eventually uniformly display it as a collection of department-specific data. The application of the data mart is Complement to data warehouse applications.

After several years of efforts, Jilin Mobile Communication Co., Ltd. has successfully established a decision support business analysis system in provincial companies, and the BI system has also gradually improved. Provincial companies extract, clean, process, organize, and load relevant business data from the business system into the data warehouse, form a basic analysis data store in the data warehouse, and guide the marketing strategy of the first-level company in the city.

Problems also arise, due to the obvious regional differences in customer groups, market capacity, and profit sources of subordinate branches. It is difficult for provincial companies to generate individuality for a single regional branch through the collection and analysis of branch data across the province. Decision support. On the other hand, in the process of developing terminal markets, prefecture-level branches have stimulated strong application requirements, which are manifested in more precise data granularity requirements, more flexible requirements, and stronger operability. .

In June 2005, China Mobile Communications Co., Ltd. formulated the "China Mobile Business Analysis System Data Mart (Pilot) Business Technical Proposal". In order to make the operation analysis system play a greater role in the daily production and operation of prefecture-level companies, Jilin Mobile finally decided to cooperate with AsiaInfo Technology to comprehensively build a "data market". Jilin City of Jilin Province became the first "crab-eating" city among 12 pilots.

Jilin Mobile hopes to timely and accurately understand the analysis needs of local companies through the construction of data marts, so as to better serve the production and marketing of first-tier local companies. Jilin City Branch also hopes to improve its own level of business analysis and implement the refined marketing strategy of the group company.

In terms of overall design, Jilin Mobile hopes to build a unified data mart model for other branches in Jilin Province through a pilot project in Jilin City, which basically covers the needs of fixed statistical reports and analysis in the cities, unified modeling, and unified management. In terms of functions, it provides individual customer analysis for the marketing behavior of the city branch, improves the implementability of the results of business analysis, supports refined marketing, and supports the development of flexible thematic analysis of cities. Develop a standardized and open data platform to meet the needs of more personalized and temporary analysis of branches in different provinces and cities in the province.

In general, Jilin Mobile put forward a very practical business description of AsiaInfo Technology, which is "based on providing rich data, providing brief analysis functions, and improving daily analysis capabilities as the main means, with the ultimate goal of solving various business goals. To vigorously improve the comprehensive data utilization and analysis capabilities of prefecture-level company data, and vigorously improve the effectiveness of active service and active marketing of branches. "

The data mart project began to organize demand surveys in June 2005. After five months of construction time, it was put into use at the end of November 2005. It completed all basic set functions required by China Mobile Group's pilot project and met local Jilin requirements. Featured expansion set content.

As the implementing party, AsiaInfo Technology has followed the principle of "platform standardization and business personalization" during the construction of the Jilin Data Mart. On the one hand, AsiaInfo uses standard system software on the data mart foundation platform to make the logical data model of the data mart unified and standard; on the other hand, when the city branch company develops application functions, it combines local actual conditions and reflects the local Demand characteristics. During the project construction period, Jilin Mobile reported to the leaders of China Mobile Group Corporation twice about the project's construction methods and ideas. AsiaInfo's construction ideas and construction results were highly recognized by China Mobile Corporation.

With the successful trial of "data mart" projects by companies such as Jilin Mobile and Yunnan Mobile, hundreds of prefecture-level companies in 31 provinces of China Mobile will launch data mart projects. It is foreseeable that 2006 will be a year for mobile companies to further explore BI applications and improve the level of BI construction. As a special data collection and analysis system, data marts will produce daily operations and management of China Mobile city-level branches This important role has become a key project for China Mobile to implement a refined business strategy.