Abstract:

Customers feedbacks are important for businesses to provide the best service and to promote their product. In this digital era, people are expressing emotions and opinions via social media towards various issues including the satisfactions of using certain products or services. One of the businesses that requires customers’ feedbacks is the telecommunication companies. Since telecommunications companies are competitive, they tend to focus more on which is the best service they can give to consumers, rather than concentrating on what consumers need. In addition, customers require other user reviews to help them choose the most suitable service based on their criteria. This work focusses on obtaining the positive and negative feedbacks by the customers using sentiment analysis method to visualize the data in an informative way. Lexicon-based approach is used during pre-processing phase as it is more suitable to manage abbreviations, symbols and language other than English. Then, Naïve Bayes classifier is applied to classify the positive and negative words. The result is represented in a graphical representation of pie chart and word cloud using Highchart as the data visualization tool. Three top telecommunication providers in Malaysia (Maxis, Celcom and Digi) are used as the case study. From the results obtained, Maxis has the most positive reviews by the customers. In order to focus on the problems faced by their customers, word cloud is used to observe the most common and important keywords used by them. The system is beneficial for businesses and customers to analyse the important features needed to meet their requirements to improve the business management as well as increase customers’ satisfaction.