Authors : Parshant; Dr. Nirmal Kaur

Volume/Issue : Volume 11 - 2026, Issue 4 - April

Google Scholar : https://tinyurl.com/7453ax3u

Scribd : https://tinyurl.com/mun6xvyn

DOI : https://doi.org/10.38124/ijisrt/26apr1276

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : Artificial intelligence assistants are widely used in domains such as customer service, education, and healthcare. However, most existing systems lack the ability to understand and respond to human emotions, which often makes interactions feel mechanical and less engaging over time. This paper presents the design and evaluation of an emotionally intelligent AI assistant framework that detects user emotions in real time and adapts its responses accordingly. The proposed system integrates a hybrid BERT-BiLSTM model trained on the GoEmotions dataset to classify user emotions into multiple categories, followed by an emotion style mapping layer that adjusts the tone of responses. A working prototype was developed using a FastAPI backend, a Next.js interface, and a SQLite database for session tracking. The system was evaluated through a user study conducted over a period of two weeks, where participants interacted with both the proposed model and a standard AI assistant. The results indicate that users showed higher engagement, increased interaction duration, and improved satisfaction when using the emotionally adaptive system. The study further highlights that emotional responsiveness plays a significant role in enhancing human-AI interaction. Additionally, the Affective Engagement Mediator model is introduced to explain how adaptive emotional behavior contributes to sustained user engagement.

Keywords : Emotionally Intelligent AI; Affective Computing; BERT-BiLSTM; Human-AI Interaction; Emotion Detection; Adaptive Response Generation; User Engagement; Affective Engagement Mediator Model.

References :

1. R. W. Picard, Affective Computing, Reprint ed. MIT Press, Cambridge, MA, USA, 2020.
2. D. Goleman, Emotional Intelligence in the Age of Artificial Intelligence. Harvard Business Review Press, Boston, MA, USA, 2021.
3. A. Folstad, C. B. Nordheim, and C. A. Bjorkli, "The role of socio-emotional attributes in enhancing human-AI collaboration," BMC Public Health, vol. 11, no. 5, 2023. https://doi.org/10.1186/s12889-023-15421-8
4. M. Gerlich, A. Pelzeter, and H. Schick, "A new research model for AI-based well-being chatbot engagement: Survey study," JMIR Human Factors, vol. 11, no. 1, e48007, 2024. https://doi.org/10.2196/48007
5. R. Sun, Y. Liu, M. Zhang, and T. Chen, "Intelligent emotion sensing using BERT-BiLSTM and generative AI for proactive customer care," Journal of Cloud Computing, vol. 12, no. 4, pp. 88-103, 2025. https://doi.org/10.1007/s10586-025-04108-3
6. Y. Zhang, J. Du, S. Ma, H. Yin, and Z. Zhao, "Emotion recognition using multimodal data and machine learning techniques: A tutorial and review," Information Fusion, vol. 54, pp. 17-36, 2020. https://doi.org/10.1016/j.inffus.2019.06.011
7. S. Thompson, "Bridging the affective gap: A decade of advancement in emotion recognition," Journal of Computer-Mediated Communication, vol. 26, no. 3, pp. 155-172, 2021. https://doi.org/10.1093/jcmc/zmab007
8. E. K. Isabirye, "Including affective computing in user experience design for emotion-aware systems," Famous Journal of Computer Science and Technology, vol. 2, no. 7, pp. 45-62, 2025.
9. I. E. Kezron, "Ethical implications of emotionally intelligent AI: Transparency and user autonomy," Cybersecurity and Smart Cities Framework, vol. 3, no. 2, pp. 110-127, 2025.
10. J. Iskef, "The influence of conversational AI on consumer behavior: A systematic review," ResearchGate Preprints, 2022. https://doi.org/10.13140/RG.2.2.27165.44003
11. H. Zhang and DeepSeek Team, "Emotion and intention detection in large language models," MDPI Mathematics, vol. 13, no. 23, p. 3812, 2025. https://doi.org/10.3390/math13233812
12. J. Devlin, M. W. Chang, K. Lee, and K. Toutanova, "BERT: Pre-training of deep bidirectional transformers for language understanding," in Proc. NAACL-HLT, Minneapolis, MN, USA, 2019, pp. 4171-4186.
13. D. Demszky et al., "GoEmotions: A dataset of fine-grained emotions," in Proc. 58th Annual Meeting of the ACL, Online, 2020, pp. 4040-4054.
14. R. Zhang, M. Li, J. Zhang, and W. Xu, "Efficient dialogue state tracking by selectively overwriting memory," in Proc. 58th Annual Meeting of the ACL, Online, 2020, pp. 1-10.
15. S. Roller et al., "Recipes for building an open-domain chatbot," in Proc. 16th Conf. EACL, Online, 2021, pp. 300-325.