Analysing Thai Social Media Content to Improve Customer Satisfaction

Abstract

The emergence of social media in Thailand has given millions of users a platform to express and share their opinions about products and services, and so social media platforms are considered to be a rich source of information for companies to understand their customers. This offers companies a fast and effective way to monitor public opinions on their brands, products, services, etc. The Metropolitan Electricity Authority (MEA) is concerned about this situation and developed a voice of the customer management system (VOCMS) that uses and analyses data related to customers, payments, electrical usage, power outage events, complaints and customer calls. In addition, it also retrieves feedback from customers via various social media channels. However, sentiment analysis performed on Thai social media has challenges brought about by language-related issues, such as the differences in writing systems between Thai and English, short-length messages, slang words, and word usage variation. This paper focuses on social media content classification and on solving data sparsity issues. We use lexicon-based techniques to classify them into positive, negative, or neutral sentiments. The procedure of analysing Thai social media content is subdivided into three modules: (1) data retrieval, (2) data pre-processing and (3) data classification. The results from this initiative project can help the MEA to improve its customer services and satisfaction by enabling quick responses to customer complaints by obtaining more details to help solve problems. Moreover, the information from analysing social media information could give customers knowledge and could be helpful for both long-term and short-term planning.