Issues in a transactional database emanating from data warehouse project

摘要

When a data warehouse (DW) is built, typical extraction, transformation, and loading operations are made, including cleaning tasks of the data, conflicts resolution, inconsistencies identification, and constrains checking, among others. In this work, two databases (DB) were combined to obtain indexes approved and failing course in basic sciences area (Mathematics, Physics, Chemistry) of the Institute of Technology of San Luis Potosí (ITSLP for their initials in Spanish). The database table intermediate of the transactional database was used to calculate dimensions and fact database tables of one warehouse. Stored data in DW are the input to Data Mining Algorithms and validation information process. This benefit was not expected. At last, the combined database was divided into a transactional database and historical database, to avoid redundancy, inconsistency, incongruity, and other integrity issues and to prepare the historical database to update DW. This DW will be suitable to carry out statistical queries in diverse areas and faculties. The final divisions of mixed database and alternative uses of intermediate tables are some non-awaiting results of original planning of project.