ItemConference Paper

AI that Understands Us: LLM-Based Emotion and Stress Insights from Online Communication

Title
AI that Understands Us: LLM-Based Emotion and Stress Insights from Online Communication
Creator
Rajak, Akash; Mishra, Siddheshwari Dutt; Vidushi; Bhardwaj, Shashank; Kumar, Amit; Kumar, Sunil
Description
The large language models (LLMs) show more accurate interpretation of human emotion and psychology from text which is available on the social media chats. This study analyses emotion recognition and stress detection capabilities of fine-tuned LLMs (GPT-2, FLAN-T5 and LLaMA-7B) using text data from social media and conversation logs. The models were tested for performance using the publicly available emotion language datasets, DailyDialog, EmotionX and GoEmotions. The models were evaluated for performance and efficiency using classification accuracy, macro-F1 score, and inference time. The results identify the performance spectrum of the models and the large models' enhanced ability to recognize detailed emotional states. These results offer real world applicability of LLM methodologies to stress detecting and decision-support automated systems in mental healthcare. 2026 IEEE.
Source
International Conference on Innovative Practices in Technology and Management, ICIPTM 2026;
Date
01-01-2026
Publisher
Institute of Electrical and Electronics Engineers Inc.
Subject
Classification; Emotion; Healthcare; LLM; Social Chat; Social Media; Stress
Coverage
Rajak A., Krishna Institute of Engineering & Technology (KIET), Uttar Pradesh, Ghaziabad, India; Mishra S.D., School of Computer Science and Engineering, Galgotias University, Greater Noida, India; Vidushi, School of Sciences, Christ (Deemed To Be University), Karnataka, Bengaluru, India; Bhardwaj S., Krishna Institute of Engineering & Technology (KIET), Uttar Pradesh, Ghaziabad, India; Kumar A., Krishna Institute of Engineering & Technology (KIET), Uttar Pradesh, Ghaziabad, India; Kumar S., Krishna Institute of Engineering & Technology (KIET), Uttar Pradesh, Ghaziabad, India
Rights
Restricted Access; Hardcopy may be available in the library
Relation
ISBN: 979-831954328-8;
Format
online
Language
English
Type
Conference paper

Citation

Rajak, Akash; Mishra, Siddheshwari Dutt; Vidushi; Bhardwaj, Shashank; Kumar, Amit; Kumar, Sunil, “AI that Understands Us: LLM-Based Emotion and Stress Insights from Online Communication,” CHRIST (Deemed To Be University) Institutional Repository, accessed June 19, 2026, https://archives.christuniversity.in/items/show/26042.

Output Formats