Multi-Modal Emotion Classification in Hinglish Text Using Neuro-Symbolic Deep Learning

The dataset comprises a comprehensive collection of multimodal Hinglish tweets annotated across a spectrum of emotions including anger, fear, happiness, sadness, frustration, compassion, mixed emotions, and others. Each emotion category, such as fear, is represented by distinct datasets: fear_emojis_dataset.csv contains tweets annotated with fear-related emojis, fear_words_dataset.csv includes tweets annotated with fear-evoking words, and fear_words_meta.json provides supplementary metadata or annotations specific to fear expressions in Hinglish tweets. This structured approach allows researchers to explore and analyze emotional expressions within Hinglish social media content, spanning both textual and visual modalities. Such datasets are pivotal for advancing research in sentiment analysis, emotion detection, and cultural studies within the context of multilingual and multicultural online interactions.