EmoSyn: Adaptive Emotion Framework for Sentiment Analysis and Internet of Medical Things

Area of Science:

• Healthcare technology
• Artificial Intelligence
• Internet of Medical Things (IoMT)

Background:

• Integrating sentiment analysis into IoMT healthcare systems requires adaptive, high-precision frameworks.
• Existing methods often suffer from latency or accuracy issues, limiting effectiveness in dynamic scenarios.

Purpose of the Study:

• To propose EmoSyn, a novel framework addressing limitations in current IoMT sentiment analysis.
• To enhance real-time IoMT system adaptability and precision through sentiment interpretation.

Main Methods:

• EmoSyn integrates Emotion Wave Modulation (EWM), Neuro-Cognitive Language Dynamics (NCLD), and Sentient IoMT Interaction Protocol (SIP).
• EWM uses high-dimensional feature vectors and kernelized mappings for emotional waveforms.
• NCLD employs synthetic neural mappings and adaptive linguistic modeling for semantic transitions.
• SIP enables real-time IoMT recalibration via bidirectional sentiment feedback.
• An Emotion-Aware Predictive Synthesis Algorithm (EAPSA) was implemented.

Main Results:

• EmoSyn achieved 91% precision on the MOSEI dataset.
• EmoSyn achieved 86% precision on the MIMIC-III dataset.
• The framework demonstrated low average latencies of 29 ms (MOSEI) and 27 ms (MIMIC-III).

Conclusions:

• EmoSyn offers a precise and efficient solution for sentiment analysis in IoMT healthcare.
• The framework effectively balances accuracy and low latency in dynamic environments.
• EmoSyn shows significant potential for improving IoMT-based healthcare applications.