Application of Data mining Techniques for student success and failure prediction

Muluken Alemu Yehuala
Information Technology, Adama University
March, 2013
Full text (external site)


This research work has investigated the potential applicability of data mining technology to predict student success and failure cases based on data sets gathered from Debre Markos University registrar office. CRISP-DM (Cross Industry Standard Process for Data mining) is a data mining methodology to be used by the research. Classification and prediction data mining functionality are used to extract hidden patterns from students data. These patterns can be seen in relation to different variables in the students records. The classification rule generation process is based on the decision tree and Bayes as a classification technique and the generated rules were studied and evaluated. Data collected from MS_EXCEL files, and it has been preprocessed for model building. Models were built and tested by using a sample data set of 11,873 records of first year-first semester regular undergraduate students enrolled in 6 schools from 1999E.c 2004 E.c. Analysis is done by using WEKA 3.7 application software. The research results offer a helpful and constructive recommendations to the academic planners in universities of learning to enhance their decision making process. This will also aid in the curriculum structure and modification in order to improve students academic performance. Students able to decide about their field of study before they are enrolled in specific field of study based on the previous experience taken from the research-findings. The research findings indicated that secondary school leaving certificate result, Sex, Number of students in a class, number of courses given in a semester, and field of study are the major factors affecting the student performances. So, on the bases of the research findings the level of student success will increase and it is possible to prevent educational institutions from serious financial strains.