Predicting employability of students using data mining approach

Educational data mining (EDM) is an emerging discipline that emphasizes on the application of data mining tools and techniques on educational data. The discipline focuses on extracting and analyzing educational data to develop models for improving learning experiences and institutional effectiveness. If this technology is made use for the benefit of the common man, then the purpose is served. The purpose of this paper is to predict employability of students using data mining approach. The study will also help the prospective engineering students in selecting or choosing a right course namely, computer science, civil, electronics etc., based on the entrance exam ranking for admission to BE course. In this paper, a combination of statistical and data miming approach is applied. Statistical techniques are used in the preprocessing stage and data mining approach in the form of algorithms are applied to solve placement chance prediction problem. Two clustering algorithms viz., X-Means and Support vector Clustering and a classification algorithm Naïve Bayes are applied on the same data set. These algorithms are implemented, to predict accurately one among the various courses offered that predict better placement chances. Student will enter Rank, Gender, Category and Sector and the model will give answer in terms of Excellent (E), Good (G), Average (A) and Poor (P) for the data entered. Each and every course offered is associated with one of the above answers viz., E, G, A, P Such as, computer science with – E, electrical and electronics with – P and so on. Algorithms applied are compared in terms of precision, accuracy and truth positive rate. From the results obtained it is found that the cluster algorithm viz., X-Means predicts better in comparison with other algorithms. This work will help the students in selecting a best course suitable for them which ensure best placement chances based on the data entered.