Miha Batič (2008) Employing data mining to transform sales information system into business intelligence. EngD thesis.
These days companies have complex information systems with a vast amount of data. Amount of data increases every day. With insufficient knowledge of data, stored in company databases, the data is worthless. In this diploma I have divided data-mining algorithms into operations and described them. During the description of these algorithms and groups of algorithms I have encountered several different classifications. From among them I have chosen the most suitable one. I focused on the usage of data-mining techniques in the sales department, which is one of the basic business functions. Everybody knows sales function; we come across the sales function every day. I described possible use of data-mining in the marketing field, risk management and fraud management. I focused mainly on marketing part since it is very important activity. In the last part of this diploma I realized some of the potential uses of data-mining in Microsoft Analysis Services. I realized: - Cross-sales, Market basket analysis, - Segmentation, - Sales forecast, - Customer retention. The data from Microsoft Dynamics NAV ERP system allowed me to perform data-mining. I also discovered, that even though Microsoft Analysis Services gives us the opportunity to perform data-mining only with connecting to relational databases, we still need data warehouse in some cases. This is very much dependent on data organization in data tables. Due to a complexity and interesting details about sales function I believe that discussing sales function is very extensive. This is also the reason why I limited by only implementing possible uses of data-mining to the above four.
Actions (login required)