ANALISIS SENTIMEN PADA KOLOM SARAN INDEKS KEPUASAN MAHASISWA TERHADAP KINERJA DOSEN

Abstract

Satisfaction from students can affect performance, because if students are not satisfied then there is something lacking with the performance of a lecturer. The purpose of this study is to determine student satisfaction with lecturer performance using sentiment analysis method. Using descriptive quantitative methods, the population of this study is active RMIK students, both regular and parallel students who have filled out a questionnaire at the end of each semester totalling 2,436 comments on the IKMKD suggestion column for 3 academic years, namely from 2020-2023. Student satisfaction with lecturer performance for positive sentiment the percentage decreased from 85% to 81%, for neutral sentiment remained 6%, and for negative sentiment the percentage increased from 9% to 13%. For neutral sentiments with comments “enough” and “no suggestions,” and for negative sentiments related to offline lectures, lack of explanation, and tight deadlines, affecting student interaction, understanding, and time. The period in this study is from 2020-2023. Samples taken directly from active students in the Medical Records and Health Information program. Using Azure Machine Learning Application to process data. The percentage of student satisfaction with lecturer performance for positive sentiment in the 2020-2021 academic year is 85% (943 comments), the 2021-2022 academic year is 82% (553 comments), and the 2022-2023 academic year is 81% (538 comments).