Analýza a návrh řešení reportingu v oblasti maloobchodního prodeje

Abstract

The thesis focuses on analysis and design of retail sale reporting solution through the use of cluster analysis method. This document is divided into three parts. In theoretical part are explained basic concepts of a Business Intelligence system and described a process of gaining knowledge from a database whose method of cluster analysis, in particular k-means, is applied for a data analysis of retail chain. In analytical section are revealed the benefits and project requirements including the user's needs. In practical part is suggested improvement of the current data model concerning new dimensional tables and fact tables. Moreover, in this part is also enlightened a process of using the data pumps to fill data warehouse, which leads to a segmentation of customers in R tool by applying the k-means method. The knowledge of a customer segmentation gained through the suggested reports is interpreted on obtained sample of data.