Proposing Malay Sarcasm Detection on Social Media Services: A Machine Learning Approach

Abstract

User comments from social media platforms have become crucial inputs for organisations, especially the government, to get feedback about their programs and services. However, since people can respond freely on social media sites, sometimes they like to use sarcastic texts implicitly in conveying their disagreeing views. Research on sarcasm detection for other languages such as the Malay language is still in its early stages. The use of noisy text, mixed languages and slang words by social media users has increased the difficulty of classifying sentiments in the Malay language. Thus, this paper proposes a Malay sarcasm detection model on social media based on a machine learning approach. The proposed model will also leverage the emotion reaction button of the Facebook platform as one of the main features to be used in sarcasm detection.