Educational Data Mining (EDM) is an evolving field exploring pedagogical data by applying different machine learning techniques/tools. It can be considered as interdisciplinary research field which provides intrinsic knowledge of teaching and learning process for effective education. The main objective of any educational institution is to provide quality education to its students. One way to achieve highest level of quality in higher education system is by discovering knowledge that predicts teachers’ performance. This study presents an efficient system model for evaluation and prediction of teachers’ performance in higher institutions of learning using data mining technologies. To achieve the objectives of this work, a two-layered classifier system was designed; it consists of an Artificial Neural Network (ANN) and Decision Tree. The classifier system was tested successfully using case study data from a Nigerian University in the South West of Nigeria. The data consists of academic qualifications for teachers as well as their experiences and grades of students in courses they taught among others. The attribute selected were evaluated using two feature selection methods in order to get a subset of the attributes that would make for a compact and accurate predictive model. The WEKA machine learning tool was used for the mining. The results show that, among the six attributes used, Working Experience, and Rank are rated the best two attributes that contributed mostly to the performance of teachers in this study. Also, considering the time taken to build the models and performance accuracy level, C4.5 decision tree outperformed the other two algorithms (ID3 and MLP) with good performance of 83.5% accuracy level and acceptable kappa statistics of 0.743. It does mean that C4.5 decision tree is best algorithm suitable for predicting teachers’ performance in relation to the other two algorithms in this work.

Educational Data Mining Decision Tree Artificial Neural Networks Machine learning WEKA