Integrating Emotion Detection with Sentiment Analysis for Enhanced Text Interpretation
Arpan Ghosh, Naimish Pandey, C. Ashokkumar · 2024
The analysis of public sentiment on various social media platforms has become more important.. However, conventional sentiment analysis often struggle to capture intricate emotional nuances present in textual data.. This study proposes an innovative approach by integrating emotion detection with sentiment analysis to improve the interpretation of text.. A novel algorithm "BERT-DeepEmo" is proposed, combining the strong contextual modeling capabilities of Bidirectional Encoder Representations from Transformers (BERT) with a deep learning-based emotion recognition model (DeepEmo). Simulation analyses were performed to evaluate the effectiveness of the proposed algorithm compared to established approaches like BiLSTM-ECNN and RoBERTa-EDTA. Evaluation metrics including accuracy, precision, and mean absolute error were employed for comparison purposes. The results confirmed that BERT-DeepEmo enhances the overall quality of sentiment analysis while considering the emotional nuances in text. These findings highlight the proposed approach's potential to offer a more holistic and effective sentiment analysis framework, thereby facilitating informed decision-making and deeper insights across various applications. Future research work may explore into further integration with alternate process of emotional data, such as speech or visual inputs, to broaden the framework's applicability and efficacy.