Analisis Sentimen Perbaikan Kualitas Layanan Indriver Menggunakan Text Mining

Abstract

Indriver is one of the companies engaged in online transportation services in the form of ridesharing with tens of millions of downloads on PlayStore. The last few reviews on the Indriver application show that the application has received a lot of negative sentiment such as application errors, inaccurate maps, difficulty finding the nearest driver. Therefore, it is necessary to conduct sentiment analysis of user reviews so that the priority factors that influence complaints are known and suggestions for improvements can be made.This study uses data on user reviews of the Indriver application in December 2022 on Google Play. The data was collected using Google Collabs and sentiment analysis was performed using text mining methods, and using worldcloud to display words that are often discussed in the reviews. Based on sentiment analysis, it is found that there are 17,229 reviews with positive sentiment and 15,495 reviews with negative sentiment. Factors that influence negative sentiment are unreasonable price negotiations, inaccurate maps, difficulty accessing features, difficulty communicating between passengers and drivers by telephone. The proposed improvements obtained are to carry out technical improvements and periodic updates, understanding the system, optimizing the ordering system, setting a minimum tariff