Emotion dataset from Indonesian public opinion

Area of Science:

• Natural Language Processing
• Computational Linguistics
• Social Media Analysis

Background:

• Public opinions and emotional responses are abundant on social media platforms like Twitter.
• Indonesia's large and active Twitter user base presents a significant opportunity for opinion data mining.
• Existing emotion text datasets in Indonesian are limited, highlighting a research gap.

Purpose of the Study:

• To develop a comprehensive dataset of Indonesian-language tweets for emotion classification.
• To facilitate research in understanding and analyzing public sentiment and emotions expressed in Indonesian text.
• To address the scarcity of labeled emotion datasets for the Indonesian language.

Main Methods:

• Collected Indonesian tweets from Twitter to form a dataset of public opinions.
• Pre-processed the gathered data through cleaning and normalization for subsequent analysis.
• Annotated the dataset with six distinct emotional labels: anger, fear, joy, love, sad, and neutral.

Main Results:

• Successfully created a novel dataset of Indonesian tweets focused on public opinion and emotions.
• The dataset is prepared for the task of classifying emotions in Indonesian text.
• The annotated data provides a valuable resource for computational emotion analysis in Indonesian.

Conclusions:

• The developed Indonesian emotion dataset fills a critical gap in available resources.
• This dataset will enable advancements in sentiment analysis and emotion detection for the Indonesian language.
• The research supports the potential of social media data mining for understanding public emotional responses.