Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation

Empathetic and coherent responses are critical in automated chatbot-facilitated psychotherapy. This study addresses the challenge of enhancing the emotional and contextual understanding of large language models (LLMs) in psychiatric applications. We introduce Emotion-Aware Embedding Fusion, a novel...

Full description

Saved in:
Bibliographic Details
Main Authors: Abdur Rasool, Muhammad Irfan Shahzad, Hafsa Aslam, Vincent Chan, Muhammad Ali Arshad
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:AI
Subjects:
Online Access:https://www.mdpi.com/2673-2688/6/3/56
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850093692644229120
author Abdur Rasool
Muhammad Irfan Shahzad
Hafsa Aslam
Vincent Chan
Muhammad Ali Arshad
author_facet Abdur Rasool
Muhammad Irfan Shahzad
Hafsa Aslam
Vincent Chan
Muhammad Ali Arshad
author_sort Abdur Rasool
collection DOAJ
description Empathetic and coherent responses are critical in automated chatbot-facilitated psychotherapy. This study addresses the challenge of enhancing the emotional and contextual understanding of large language models (LLMs) in psychiatric applications. We introduce Emotion-Aware Embedding Fusion, a novel framework integrating hierarchical fusion and attention mechanisms to prioritize semantic and emotional features in therapy transcripts. Our approach combines multiple emotion lexicons, including NRC Emotion Lexicon, VADER, WordNet, and SentiWordNet, with state-of-the-art LLMs such as Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4. Therapy session transcripts, comprising over 2000 samples, are segmented into hierarchical levels (word, sentence, and session) using neural networks, while hierarchical fusion combines these features with pooling techniques to refine emotional representations. Attention mechanisms, including multi-head self-attention and cross-attention, further prioritize emotional and contextual features, enabling the temporal modeling of emotional shifts across sessions. The processed embeddings, computed using BERT, GPT-3, and RoBERTa, are stored in the Facebook AI similarity search vector database, which enables efficient similarity search and clustering across dense vector spaces. Upon user queries, relevant segments are retrieved and provided as context to LLMs, enhancing their ability to generate empathetic and contextually relevant responses. The proposed framework is evaluated across multiple practical use cases to demonstrate real-world applicability, including AI-driven therapy chatbots. The system can be integrated into existing mental health platforms to generate personalized responses based on retrieved therapy session data. The experimental results show that our framework enhances empathy, coherence, informativeness, and fluency, surpassing baseline models while improving LLMs’ emotional intelligence and contextual adaptability for psychotherapy.
format Article
id doaj-art-7bdddcc5de654a6797a3a639f5858062
institution DOAJ
issn 2673-2688
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series AI
spelling doaj-art-7bdddcc5de654a6797a3a639f58580622025-08-20T02:41:51ZengMDPI AGAI2673-26882025-03-01635610.3390/ai6030056Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response GenerationAbdur Rasool0Muhammad Irfan Shahzad1Hafsa Aslam2Vincent Chan3Muhammad Ali Arshad4Department of Information and Computer Sciences, University of Hawaii at Manoa, Honolulu, HI 96822, USASelTeq, University Avenue, Palo Alto, CA 94301, USASchool of Computer Science and Technology, Donghua University, Shanghai 201620, ChinaDepartment of Information and Computer Sciences, University of Hawaii at Manoa, Honolulu, HI 96822, USADepartment of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, ChinaEmpathetic and coherent responses are critical in automated chatbot-facilitated psychotherapy. This study addresses the challenge of enhancing the emotional and contextual understanding of large language models (LLMs) in psychiatric applications. We introduce Emotion-Aware Embedding Fusion, a novel framework integrating hierarchical fusion and attention mechanisms to prioritize semantic and emotional features in therapy transcripts. Our approach combines multiple emotion lexicons, including NRC Emotion Lexicon, VADER, WordNet, and SentiWordNet, with state-of-the-art LLMs such as Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4. Therapy session transcripts, comprising over 2000 samples, are segmented into hierarchical levels (word, sentence, and session) using neural networks, while hierarchical fusion combines these features with pooling techniques to refine emotional representations. Attention mechanisms, including multi-head self-attention and cross-attention, further prioritize emotional and contextual features, enabling the temporal modeling of emotional shifts across sessions. The processed embeddings, computed using BERT, GPT-3, and RoBERTa, are stored in the Facebook AI similarity search vector database, which enables efficient similarity search and clustering across dense vector spaces. Upon user queries, relevant segments are retrieved and provided as context to LLMs, enhancing their ability to generate empathetic and contextually relevant responses. The proposed framework is evaluated across multiple practical use cases to demonstrate real-world applicability, including AI-driven therapy chatbots. The system can be integrated into existing mental health platforms to generate personalized responses based on retrieved therapy session data. The experimental results show that our framework enhances empathy, coherence, informativeness, and fluency, surpassing baseline models while improving LLMs’ emotional intelligence and contextual adaptability for psychotherapy.https://www.mdpi.com/2673-2688/6/3/56large language modelspsychotherapy chatbotsemotion-aware embeddinghierarchical fusionattention mechanismsemotion lexicon
spellingShingle Abdur Rasool
Muhammad Irfan Shahzad
Hafsa Aslam
Vincent Chan
Muhammad Ali Arshad
Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
AI
large language models
psychotherapy chatbots
emotion-aware embedding
hierarchical fusion
attention mechanisms
emotion lexicon
title Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
title_full Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
title_fullStr Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
title_full_unstemmed Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
title_short Emotion-Aware Embedding Fusion in Large Language Models (Flan-T5, Llama 2, DeepSeek-R1, and ChatGPT 4) for Intelligent Response Generation
title_sort emotion aware embedding fusion in large language models flan t5 llama 2 deepseek r1 and chatgpt 4 for intelligent response generation
topic large language models
psychotherapy chatbots
emotion-aware embedding
hierarchical fusion
attention mechanisms
emotion lexicon
url https://www.mdpi.com/2673-2688/6/3/56
work_keys_str_mv AT abdurrasool emotionawareembeddingfusioninlargelanguagemodelsflant5llama2deepseekr1andchatgpt4forintelligentresponsegeneration
AT muhammadirfanshahzad emotionawareembeddingfusioninlargelanguagemodelsflant5llama2deepseekr1andchatgpt4forintelligentresponsegeneration
AT hafsaaslam emotionawareembeddingfusioninlargelanguagemodelsflant5llama2deepseekr1andchatgpt4forintelligentresponsegeneration
AT vincentchan emotionawareembeddingfusioninlargelanguagemodelsflant5llama2deepseekr1andchatgpt4forintelligentresponsegeneration
AT muhammadaliarshad emotionawareembeddingfusioninlargelanguagemodelsflant5llama2deepseekr1andchatgpt4forintelligentresponsegeneration