Work with Free Response Implementation Process Analysis Based on Sequential Pattern Mining in Engineering Education

Abstract

Learning process, its dynamics are one of the points of focus of researchers which conduct studies in the field of education. The widespread use of learning management systems and MOOC platforms allows to create and store records of student activities in educational system (logs). In educational process special systems for implementing assignments of a particular type, for example, software development, construction of concept maps, can be used together with MOOC platforms or learning management systems. These special systems can also keep logs of student activities. The logs are an analytical issue in applying educational data mining, in particular in sequential pattern mining. The use of sequential pattern mining makes it possible to identify typical patterns of student activities during the learning process. These patterns may be used in defining potential problems for students and courses with a goal of improving educational process through course corrections or through providing special feedback to students with problems. However, the usage of sequential pattern mining approach is limited to the majority of works with free response, widely applied in engineering education. A method of work with free response implementation process analysis based on sequential pattern mining is proposed in the paper. The method was tested in training future engineers in the field of information technologies in Ufa State Aviation Technical University on the results of distance term paper preparation in the course "The design of automated systems for information processing and control". The use of the suggested method allows to extract undesirable patterns of implementing works with free response and reasons for their presence. Also, the method allows to define typical strategies of implementing such works and their proportions.