Mining Student's Reviews to Obtain Their Perception toward College Department Performance

Yuliska Yuliska (Politeknik Caltex Riau)
Dini Hidayatul Qudsi (Politeknik Caltex Riau)
Lya Anggraini (Institut Teknologi Bandung)
Juanda Hakim Lubis (Universitas Medan Area)

Proceeding International Applied Business and Engineering Conference
2021: Proceeding International Applied Business and Engineering Conference 2021

Student’s perception toward department performance can be crucial, therefore, it can be used to evaluate the department outcome and take an immediate action to improve its management. This study applies sentiment analysis and topic modeling to the student’s reviews of college department at Politeknik Caltex Riau in order to mine student’s perception for seven college departments performance. Sentiment analysis with Support Vector Machine (SVM) is employed to obtain student’s sentiment. There are 3 types of sentiments to be analyzed; positive, negative and neutral. Topic modeling with Latent Dirichlet Allocation (LDA) is also carried out to get some important keywords in the student’s reviews. Our experiments show that Positive is the most prominent sentiment in the student’s reviews while LDA reveals some important topics toward preferences.