Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{203899,
        author = {swanand uttarwar and Sanyukta Deshmukh and Advait Ugavekar and Aary Upare and Vishal Wale},
        title = {Emotion and Sentiment Evolution},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {523-530},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=203899},
        abstract = {The conversation agents grow rapidly creating many limitations many agents are not that connective to the emotional nature of the humans, which affect the interaction between human and machine. In this paper we are going to study about the Artificial Intelligence - AI for affective computation. We are going to keep track of the progress of emotion detection from basic text to advanced reasoning. We are going to start with basic text-based method of conversation used in most ML models such as Naïve Bayes and Support Vector Machines (SVM), and we are going to compare them with deep learning neural networks which achieves high accuracy (for example, an F1-score of 0.95 for sadness detection). Then we are going to move toward context-based Emotion Recognition in Conversation (ERC), also going to focus on some pre-trained language models like BERT which are fine-tuned for including dialogue structure and speaker information. Along with this we are going to use the approach of Multimodal Emotion Recognition (MER), using which we can combine text, audio, and some visuals to reduce the “heterogeneity gap,” with the help of both supervised and unsupervised learning methods. At the end, we are going to discuss about the final stage of emotion evolution AI systems that understand casualty in emotions. In this we have included models which help us to identify the cause of emotion generation to generate the response which can show empathy, along with this we are trying to use Retrieval-Augmented Generation (RAG) frameworks such as “Cause Motion,” which keep the track long-sequence causal patterns in multimodal conversations. Overall, this research focuses on the change from simple emotion classification to a more complete, cause based understanding of human emotions.},
        keywords = {Emotion, Sentiment Evolution, Artificial Intelligence (AI), Natural Language Processing (NLP), Affective Computation, Detection of Emotions, Machine Learning, Neural Networks, Human-Machine Interaction, Emotion Recognition in Conversation (ERC), Multimodal Emotion Recognition (MER), Cross-Modal Fusion, Emotional Causality, Retrieval-Augmented Generation (RAG), Empathetic Response Generation},
        month = {June},
        }