Analisis Waktu dan Lokasi Klasifikasi Sentimen Twitter Terhadap Kinerja Layanan Provider Telekomunikasi Menggunakan Naïve Bayes

Abstract

Telecommunications users in Indonesia continue to grow rapidly from year to year. Along with the public's increasing need for communication, whether via SMS, telephone or data services, there is competition among telecommunications providers to attract or retain customers. Customer Opinion shows the level of service quality provided by the provider. Various opinions expressed by customers about telecommunications providers can be known through the social media Twitter. Twitter as a type of microblogging that produces raw data that can overwhelm its users, one solution to this problem is to classify raw data. The Naïve Bayes method can handle text or documents. The documents used in this research are comments from Twitter users. The Naïve Bayes method was chosen because of its ability to handle large text data with sufficient accuracy and an efficient computational process (Aggarwal & Zhai, 2012). It is hoped that the results of this analysis will provide insight for telecommunications providers in improving the quality of their services in accordance with user needs and expectations (Smith et al., 2023). This research uses techniques to categorize customer sentiment opinions towards telecommunications providers in terms of time and location in Indonesia. Sentiment analysis only includes positive, negative and neutral classes. The expected benefit in this research is that telecommunications providers can evaluate performance and services based on time and location to achieve customer satisfaction from various complaints faced, as well as build more effective communication strategies.