Transformer-Based Indonesian Language Model for Emotion Classification and Sentiment Analysis
Hendri Ahmadian, Taufik Fuadi Abidin, Hammam Riza, Kahlil Muchtar · 2023
The rapid development of social networks has made much user-generated data accessible for public evaluation. These data can be used for multiple purposes, such as textual analysis of comments and reviews. This article employs a variant of the bidirectional encoder representations from Transformer (BERT) model designed explicitly for Bahasa Indonesia (called IndoBERT model) to enhance the performance of the Indonesian natural language understanding benchmark in tasks, such as sentiment analysis and emotion classification. The two tasks were tested using a hybrid method, combining the IndoBERT model's last hidden layer summation with a neural network model. The performance of the resulting model was assessed using the F1-score metric. The experimental results show that the proposed model attains an accuracy of 0.92 and 0.76 for sentiment analysis and emotion classification, respectively.