Mining continuous assessments marks to predict final results

Abstract

Educational Data Mining is used to study the data available in the Universities, Higher Educational Institutes and other educational fields and extract the knowledge from it. As a result ofreducing the cost of processing data and storing data, data storage became more easy and cheaper. Education Institutions are facing important and fast growth ofthe volume of educational data. Data mining also called as Knowledge Discovery in Database (KDD) and search for inter relationships and patterns that can find, but already hidden among the vast volume of educational data. Classification methods like decision trees, rule mining, Bayesian network etc can be applied on the educational data for predicting the students performance in examinations. This prediction will help the lecturers, teachers, tutors and students themselves to identify students’ performance in the end semester examination. It will help the intelligent students to motivate more to maintain higher standard ofmarks and motivate weak students score better marks. The J48 decision tree algorithm is applied on students’ internal assessment marks to predict the grade they would gain at the end semester examination. In order do more accurate prediction some personal attributes like gender, their academic district, Advanced level Stream had been considered. With this research, students’ who are likely to get higher grade or lower grade will be predicted more accurately. Predicted results can be distributed among teachers and tutors and necessary steps can be taken to improve the performance ofthe students who will be predicted to get lower grade or fail.