Decision Support Systems for Entrance Examinations using Bloom’s Taxonomy and Support Vector Machines
Admissions to post graduate courses in Karnataka, especially the MCA courses in many government colleges happen through an entrance examination. This entrance examination is being attempted by close to 7000 candidates every year. Students in the final year of degree programmes like B.Sc, BCA , B.Com are eligible to take up this examination. The work in this research paper is an effort to investigate how effectively the question paper of this examination is testing the aptitude of the potential candidates. In this direction, an attempt is being made in order to study the aptitude of the undergraduate students. Standard benchmarks have been evolved by the researchers to evaluate the aptitude of the students. The results of the candidates in the entrance examination over the past 12 years have been studied. Definite patterns of scoring in these examinations have been observed as a result of which, the following research question was has posed for the current research work. “Does the Post-Graduation examination test the aptitude of the candidates, if so, to what extent?” The research work considers the analysis of extensive sets of data sets across the results of the MC Aentrance examinations. Bloom’s taxonomy of verbs is employed to classify the questions and support vector machines are being used to identify clusters of students of similar aptitudes. Control group of students are given a variety of questions to justify the aptitide levels formed. This work can be used to generate blue prints of question papers across any of entrance examination which follow the same pattern.