Formální konceptuální analýza nad vícehodnotovými daty v systému eLogika

Abstract

The goal of this master thesis is to create a tool for computation many-valued formal concept analysis of data from e-learning system eLogika and to use this tool on this data to find connections inbetween. This data represents the behaviour of students in the system. Next, describe the theory of many-valued formal concept analysis and everything what is important to fully understand the topic, based on the mathematical relation structures perspective. The theoretical part should be used in lectures. The results of the tool show dependences between students and their evalutations in the course, which we discuss in detail. Also we compare the results, which are found by different approaches. Those two main approaches are based on statistical methods and methods based on evolution algorithms. we propose those evolution algortihms as more effective method to found relevant scalings of attributes. In addition, we have found such attributes which compared to others attributes have significantly affected evaluations of student in analysed course.