Mineração de dados em múltiplas tabelas fato de um data warehouse.

Abstract

The progress of the information technology has allowed huge amount of data to be stored. Those data, when submitted to a process of knowledge discovery, can bring interesting results. Data warehouses are repositories of high quality data. A procedure that has been adopted in big companies is the joint use of data warehouse and data mining technologies, where the process of knowledge discovery takes advantage over the high quality of the warehouse s data. When the data warehouse has information about more than one subject, it also has more than one fact table. The joint analysis of multiple fact tables can bring interesting knowledge as, for instance, the relationship between purchases and sales in a company. This research presents a technique to mine data from multiple fact tables of a data warehouse, which is a new kind of association rule mining.