Classification and Sentiment Analysis of the Facility and Service Questionnaire for Qomaruddin University Gresik

Authors

  • Muhammad Ajir Muzakki Institut Sain dan Teknologi Terpadu Surabaya
  • Gunawan Institut Sain dan Teknologi Terpadu Surabaya

Keywords:

Classification, Sentimen Analysis, Rocchio Algorithm, Naïve Bayes Algorithm

Abstract

Qomaruddin University is a private collage that always prioritizes student satisfaction as customers by providing the best and qualified services. In improving facilities and services , each semester questionnaires are conducted to obtain feedback on academic services, learning processes and campus facilities and infrastructure. The form of the questionnaire is in the form of questions with optional answers and descriptions. In the descriptive data requires further handling, considering that this descriptive data makes it very possible to reveal things that have not been revealed through question items with a limited number of optional answers. The purpose of this research is to classify the student questionnaire data using the Rocchio algorithm and sentiment analysis using the Naïve Bayes algorithm. The Rocchio algorithm is used to classify facilities and services, namely Infrastructure (Sarpras), Academic Administration Bureau (BAA), Financial Administration Bureau (BAK), Information Systems Bureau (BSI), etc. The Naïve Bayes algorithm is used for sentiment analysis, namely positive, neutral and negative. Based on the results of trials that have been carried out with 10 trials obtained, the accuracy value for the classification of facilities and services category has an average value of 83.27 while for sentiment analysis has an average value of 80.67 so it can be said that the system performance is quite stable.

Author Biographies

Muhammad Ajir Muzakki, Institut Sain dan Teknologi Terpadu Surabaya

Program Pasca Sarjana, Magister Teknologi Informasi

Gunawan, Institut Sain dan Teknologi Terpadu Surabaya

Program Pasca Sarjana, Magister Teknologi Informasi

2

Published

2020-12-31