Sentiment Analysis on the Results of the Laboratory Facilities and Infrastructure Assessment Questionnaire Using the Holistic Lexicon Based Method

Abstract

This study aims to classify sentiments from the laboratory facilities and infrastructure assessment questionnaire into three main categories, namely positive, negative, and neutral sentiments. The data used comes from a questionnaire containing opinions, comments, and suggestions about laboratory conditions. The method applied is Holistic Lexicon Based, which conducts sentiment analysis based on a sentiment dictionary to determine the orientation of each opinion word. Before the classification process, data from the questionnaire is processed through a preprocessing stage to improve the quality of the analysis results. After that, sentiment classification is carried out by assessing the orientation of the words in the opinion. The test results show that the Holistic Lexicon Based method is able to identify opinion sentences and classify sentiments with an average accuracy of 81.18%. This level of accuracy is influenced by the number of opinion sentences identified and the suitability of the sentiment dictionary used in the analysis process. This study is expected to help organizations in making decisions based on the results of sentiment analysis from the questionnaire that has been processed.