Utilization of machine learning to improve usability of the feedback process on student learning evaluation results

Agus Hermanto (Universitas 17 Agustus 1945 Surabaya)
Mochamad Sidqon (Universitas 17 Agustus 1945 Surabaya)
Nurul Fadilah (Fit Tech Inova Global)

JSI: Jurnal Sistem Informasi (E-Journal)
Vol 13, No 2 (2021)

This research aims to design an E-learning system with machine learning implementation to assist the process of evaluating student learning outcomes for improving the learning process. The method used is the Cosine Similarity Matching which is a method to measure the comparability between documents or text. In this study, there were 15 students and 15 teachers who were used as samples in the implementation of E-Learning system and evaluation of the usability of E-Learning system using the SUS method. The system testing carried out obtained a fairly good performance, namely the precision test results had a performance of 86.67% and based on the usability test results obtained a value of 61.33 which was counted in the Marginal High category. Thus the E-Learning system that has been built can help facilitate the process of evaluating student learning. It is hoped that this implementation can help make online learning easier.Keywords: Autocorrection Essay, Cosine Similarity, E-Learning, Information System, Machine Learning