Assessment of essay quiz system in the LMS Vidyanusa platform using the Nazief and Adriani Stemming Algorithms “Comparison of teacher, expert rubric, and system assessments”

This study focuses on the assessment system of the essay quiz answers on the LMS VidyaNusa platform using Nazief and Adriani algorithms. The stemming algorithm is a word cutting process that makes use of basic words using certain rules. This method uses LSA (Latent Semantic Analysis) which determines the similarity values of two documents, to find interaction between sentences using SVD (Singular Value Decomposition) algebraic calculation of each statement then to get the similarity value from each statement using cosine similarity algorithm. This system is followed to assist the teacher in evaluating answers to student essay quizzes because this system is equipped with value weighting so that it reduces subjective assessors and increases accuracy in assessment. The results of this study from 15 students by answering 10 questions and 1 teacher obtained results, that the assessment given by the teacher without the vulnerable rubric was 8-10, the value given by the vulnerable system was 0-10, and the assessment of the system with the vulnerable rubric was equal 0 -10. So when viewed from the function of the rubric on the objectivity of the assessment, the value given by the teacher without rubrics is less objective. Correlation or relationship between expert assessment system using rubric with the results of r = 0.82166624 So that both variables have a close relationship and the form of the relationship is linearly positive.