This work consists of a very brief description of eye tracking system with characteristic of all types of eye movements, which affect analysis of acquired data. These data are processed with help of various algorithms. Basic principles of algorithms are also quoted. Aim of this work is data analysis of data of eye tracking system with cluster methods. Information about cluster methods and algorithms were adopted from literature. Author uses all the acquired knowledge to test respondents. The testing used MATLAB program. Four maps were tested by two people. Attention was paid to presentation of differencies in process of making clusters. This fact is stressed more than physiological (and other) attributes of tested people. Results of testing serve to recommend the most suitable methods for geographical tasks. During creation of this study many uncertainties and problems occured. The most important ones are described. In the beginning of creating this work the first defect and difficulty in various coordinate systems, which are used by different companies dealing with eye tracking system, occured. This fact was solved by production of own programs. Despite this the question is if better interoperability like in other spheres and cooperation of all companies in this branch, will lead to satisfaction of all sides and simplifying many researches. Next topic to solve nearly relates with mentioned problematics of coordinate systems and with visualization. Due to the necessity of conversion of output clusters to pixel coordinate system the problem with their visualization arised. Visualization finally ran in GIMP program, in case of huger testing the visualization would take big amount of time. According to the character and range of work it was impossible to test all algorithms (and their parameters) on higher level. It would be desired to the future to change this aspect and to test each algorithm precisely. In terms of this paper a website for its presentation was created. In the text all the steps done during making this paper were described.