Analisis Sentimen Ulasan Customer Kopi TMLST Menggunakan Algoritma Naïve Bayes

Abstract

The rapid development of Coffeeshop is currently influenced by advances in internet technology, the existence of online food applications and websites, such as Shopeefood, and Google Maps, can help people place online orders that have no time limit. However, there are problems that arise over time such as, in collecting feedback from customers the more review data available on Google Maps and online food applications, namely Shopeefood. Therefore, a solution is needed that can help TMLST Coffee to collect, process, and analyze feedback from customers on online food applications such as Shopeefood and Google Maps in a better and more structured manner. In this study, retrieving and collecting customer review data was carried out using web scrapping techniques taken through online food applications, namely Shopeefood and Google Maps, but collecting review data was also carried out by distributing questionnaires via google forms filled out by TMLST Coffee customers. Furthermore, the method used in this research is Naïve Bayes which aims as a classification method and is able to classify customer comments into positif or negatif. And review data processing is done using the Cross-Industry Standard Process for Data Mining (CRIPS-DM) method. The CRIPS-DM stage involves the research and implementation process of the stages that have been carried out previously. The results of this study produce a high level of accuracy in predicting positif and negatif sentiment, with an accuracy of 0.82 or 82%. In addition, it produces a positif recall of 0.76 or 76% and a negatif recall of 0.89 or 89%. indicating that the model has a good ability to identify correctly. With the evaluation results of the model used, it gives an indication that Naïve Bayes can be an effective choice in conducting sentiment analysis on TMLST Coffee review data.