In today’s rapidly evolving educational landscape, fostering emotional intelligence within the classroom has become increasingly vital for promoting effective learning. Teachers often face the challenge of navigating a diverse array of student emotions, which can significantly influence engagement and comprehension. Recognising the importance of addressing emotional dynamics in educational settings, Dr Srabani Basu along with our alumni Shreyashi Shankar Sinha (B. Tech CSE- 2025) and Ms Abhishikta Datta (B. Tech CSE- 2025) have come up with a groundbreaking solution: an Emotion-Aware Conversational AI. This innovative system is designed to support educators by enhancing classroom interactions through real-time emotion detection and advanced communication strategies. By leveraging both Neuro-Linguistic Programming techniques and culturally sensitive adaptive responses, the AI aims to cultivate a more empathetic and constructive dialogue between teachers and students, ultimately enriching the learning experience and facilitating positive belief shifts.

Abstract:
This project introduces a novel Emotion-Aware Conversational AI designed to empower teachers in fostering emotionally intelligent classroom dialogue. The system uniquely integrates real-time emotion detection from textual input with Neuro-Linguistic Programming (NLP)-informed belief reframing techniques to facilitate empathetic and pedagogically effective interactions. Employing fine-tuned transformer models for nuanced emotion recognition, the AI also incorporates cultural sensitivity through adaptive response strategies guided by contextual metadata. Trained on education-specific datasets to reflect authentic classroom communication patterns, the system aims to go beyond simple emotion detection by actively guiding conversations towards constructive belief formation. Prototype testing demonstrates the AI’s ability to dynamically adjust its tone, promote empathetic exchanges, and its significant potential to deepen emotional engagement and facilitate positive belief shifts within learning environments.

Explanation of the Research in Layperson’s Terms:
Imagine a smart chatbot for teachers that reads what students write (like chat messages or journal entries) or even looks at classroom photos to spot if kids are feeling happy, confused, frustrated, or anxious. Instead of just saying “That’s okay,” it uses clever language tricks from psychology (called Neuro-Linguistic Programming) to gently reframe negative thoughts—like turning “I’m bad at math” into “You’re getting stronger at math with practice”—while being sensitive to different cultures. This helps teachers respond with empathy, keep students engaged, and create a supportive classroom where everyone feels understood and motivated to learn.

Practical Implementation and Social Implications:
The inventors have built a live web app prototype using Streamlit for an easy chat interface where teachers can input student text or upload images (e.g., handwritten notes or classroom snapshots). It runs on a hybrid setup: local AI models for offline privacy-focused use and Google Gemini API for advanced multimodal analysis. Key features include emotion mapping (e.g., “frustration” to “needs encouragement”), NLP-inspired prompts for reframing responses, and conversation history tracking.

Social Implications:
In diverse, multicultural classrooms, this AI addresses emotional gaps in education by helping teachers detect subtle stress or disengagement early, reducing dropout risks and boosting mental health. It promotes equity by adapting to cultural nuances, fostering inclusive environments where students from varied backgrounds feel validated. Long-term, it could scale to global edtech platforms, enhancing teacher well-being (by easing emotional labor) and student outcomes, with studies showing up to 89% relevance in real-time feedback.

Our Collaborations:

• Core Team: Shreyashi Shankar Sinha (Lead Developer, AI Integration) and Abhishikta Datta (UI/UX Design, Dataset Curation).
• Academic Guidance: Dr. Srabani Basu (Project Supervisor, Department of Literature and Languages, SRM University-AP) for NLP psychological insights; Prof. Murli Krishnan (Head of Department, Computer Science & Engineering) for technical oversight.
• Institutional Support: SRM University-AP’s School of Engineering & Sciences provided resources, including access to Hugging Face datasets and Google Cloud credits.
• External Influences: Drew from open-source communities (e.g., Hugging Face Transformers) and referenced works like Bandler & Grinder’s NLP foundational texts.

Our Future Research Plans: Building on this prototype, we plan to expand the emotional spectrum to include subtler states like boredom or subtle cultural pride, integrating voice analysis for hybrid (text/audio) inputs. We’ll refine calibration with larger, multilingual datasets from Indian classrooms and test in real schools via partnerships with edtech firms. Long-term, aim for a full patent on the reframing algorithm and deployment as an open-source tool to democratize emotional AI in under-resourced education systems.