Mapping Engineering Course Outcomes to Program Outcomes Using an Intelligent Framework

Abstract

Outcome Based Education is an important aspect as per National Education Policy 2020. To accomplish Program Outcomes, Course Outcomes are crucial. A group of subject matter experts typically maps Course Outcomes to Program Outcomes. This is a tedious process. To overcome this an intelligence framework is proposed in this paper. Course Outcome statements along with mapped program outcomes are considered as a data set in this paper. Split the data set as training and testing sets. A training data set is used to train the Machine Learning algorithm. Course Outcomes mapped with Program Outcomes by the expert team are considered input for the system. Machine Learning enables automated processes. Here, text documents are processed using tokenizer, removal of stop words, count words, and numeric data conversion. This numeric data is given as one input to the Machine learning algorithm. Mapped Program outcomes are given as other input to the Machine Learning algorithm. These are denoted as labels. Support Vector Machine is considered to develop intelligence for the system. A testing data set is applied to the intelligence framework to observe the performance of the system. A confusion matrix is formed based on the testing results. Traditional methods take more time to map COs to POs. Impact is more with for large curriculums. Further, it requires significant human effort. In the proposed methodology, COs to POs mapping is implemented fast and efficient manner. Further, mapping can be generated instantly with minimal human input. The accuracy of the system is calculated from the confusion matrix. Experiments indicate that Course outcomes mapped using an intelligence framework give better accuracy.