Evaluation of Teacher's Performance using Students Feedback with Pearson Correlation Coefficient and Support Vector Machine

In this research, the evaluation of teachers' comprehensive literacy using the student’s feedback has become important prediction to predict the performance of teachers. The performance of teacher is evaluated by traditional approaches such as surveys, student feedback, observations. Prior methods consume more time to evaluate performance through manual method and failed to analyze the non-linear relation with students such as student engagement, effectiveness of technical skills, teaching performance. To overcome these challenges and to evaluate the performance of teachers in linear and non-linear relations with students effectively, a Machine Learning (ML) algorithm namely Support Vector Machine (SVM) is used. Initially, data is collected from Jazan University which consists of academic performance, technical skills, examination results. Then, pre-processing is performed through min-max technique which subtracts the extracted data into a particular range. After that, feature extraction is performed to obtain processed data by using Principal Component Analysis (PCA) and for reducing dimensionality in evaluation of multiple features such as academic performance, examination results, technical skills. Finally, the classification is performed using the PCC-SVM and effectively analyzed non-linear relationship and has taken shorter time for evaluating the teacher’s performance. The proposed PCC-SVM has achieved a high accuracy of 98.71%, precision 98.89 %, F1 score of 99.54% and recall as 96.35 % when compared to existing k-Nearest Neighbors (KNN).