Analysis And Assessment Of Data Quality For Data Warehouse

The extensive application of information technology enables efficient and flexible enterprise operations, but it also brings the problem of "data explosion". A lot of useful history data are put away, and people are lost in the ocean of new data. Therefore, how to organize and store data effectively and find commercial value hidden in complicated information becomes a highly regarded question of decision makers. Data warehouse technology, which is an effective and multifunctional method to manage data, can help decision makers of all levels to estimate operation performance of enterprises and find out problems in time. It helps to settle a sound basis for managers. Currently, more and more companies begin to apply data warehouse technology.The core of data warehouse technology is "data". The quality of data is the strong support of this technology. If there are too many "issue data" in this warehouse, the information which users attain may be mistaken, and it can mislead their decisions, finally lead to immeasurable loss. So whether we have high-quality data to support data warehouse is the key point affecting the result of the warehouse project. Under this circumstance, the data quality of the data warehouse has become a heat topic among researchers in home or abroad. This article will begin a comprehensive analysis based on the researches in the past. First, according to the domestic and foreign research, the definition for data quality of data warehouse is summarized. And the necessary data quality dimensions of data warehouse are classified. Also, this paper proposes a common data quality management process and analyses its implementation. And then, a quantitative analysis module will be raised to estimate quality index of data source, and help to find what kind of data are suitable to be the value or dimension of data warehouse. These suitable data can be references when designing a warehouse. Also, this module can be used to assess the quality of data warehouse. Besides, the paper will start a deep research on the estimation and improvement of data quality. An estimation system will be raised including five elements, data storage, data set, data role, index and regulation. And an estimation module will be given based on the five elements. This module can help to define rules to assess roles of data, and mark for the data quality of relative data sets. After that, this article will introduce several in home and abroad common used methods for improving data quality, in order to improve the quality of problem data attained in data estimation. At last, some relative cases will be explained to prove the former researches.