Testing of system OptiMiner

The thesis contributes to transfer of software used in data mining lessons. The software is being transferred from academical solution LISp-Miner to the solution OptiMiner based on LOTYLDA framework that is used by real company to solve real data mining tasks. Purpose of the thesis is to create manual which describes how to work with the new solution and most importantly creation of typified tasks used in the lessons. Secondary goal of the thesis was to get system into the state where user is able to create, update, run and evaluate tasks. Evaluation of the solution was done by manual testing. Creation of documentation was based on CRISP-DM methodology and description of typified tasks. Outcome of the thesis is document that describes how to work with the new system and how to create and work with data mining tasks there. Lecturers of the data mining courses are the key audience for this document. It should help them to easily become familiar with new system and relevant parts of this manual will be citated in the new (student) study materials for the courses. Examples of solved and more importantly unsolved problems within the new system are mentioned in the conclusion. After they are solved, system should be able to cover all of the typified tasks. The thesis has two goals. Main goal was to create document describing how to work with new system OptiMiner. Most important thing was description of how to work with typified tasks used within data mining course lessons taught at Prague University of Economics and Business within new system. Secondary goal is to make System Optiminer accessible via web environment of framework LOTYLDA. This environment needed to be tested and confirm that OptiMiner is in state which allows user to work at typified tasks. At last there is mention of the list containing recommendations to be made to the official documentation of new system in conclusion.