Postup tvorby multidimenzionálního datového modelu z diagramů MMABP

This diploma thesis deals with issue of linking business process models with the area of multidimensional data models according to Business Intelligence. The main objective is to develop a methodical procedure for creating a multidimensional data model based on the indicators and dimensions found in process diagrams. This objective consists of four other sub-objectives that constitute the procedure itself. The last sub-objective validates the proposed procedure on a real customer journey of mortgage loan negotiation. The development of the procedure is based on the MMABP process methodology and Kimball's multidimensional technique. The theoretical part of the thesis offers a summary of the basic components of BI, multidimensional modelling, process modelling and the components of the MMABP process methodology used. The practical part presents a methodological framework on how to find dimensions and indicators in the created process diagrams, how to create a data structure in a relational database system environment using a Class diagram and then use this to create dimension and fact tables corresponding to the information found in the diagrams. The proposed procedure is then verified on a real customer journey as well as with an example of a possible final implementation of the multidimensional data model in the Microsoft Power BI analysis tool. The contribution of this paper is the development of a detailed methodological procedure on how to implement a multidimensional data model based on process diagrams. A partial contribution is the demonstration of the application of the procedure on the created process diagrams of the mentioned customer journey.