Abhishek Vichare | University of Mumbai (original) (raw)

Uploads

Papers by Abhishek Vichare

International Journal of Applied Engineering & Technology (London), 2023

Emotion gives thoughts and feelings of a person. Emotions are reflected from text, speech, gestur... more Emotion gives thoughts and feelings of a person. Emotions are reflected from text, speech, gestures as well facial expressions. Recognizing and interpretation of emotion from such reflections plays an essential task in the interaction between human and machine conversation. Different views or opinions in terms of sentiments are derived from such emotions. Sentiment analysis provides negative, positive, or neutral terms; whereas emotional analysis provides deeper analysis of participant's emotions that tries to drill down into the various user behaviors. Emotion Recognition and Sentiment Analysis together can help to identify, process and illuminate human affects. In this study, a diverse array of algorithms, including LSTM, BiLSTM, CNN, and RCNN, as well as ML algorithms such as RF, MNB, SVM, and Logistic Regression, were employed for the purpose of recognizing emotions from text. Models were trained and tested along with different word embedding methods and activation functions. Well balanced ISEAR dataset (International Survey On Emotion Antecedents and Reactions) is utilized in experimentation. Implementation of a fine-tuning and evaluation process for a BERT-based model using the Hugging Face Transformers library is also done for ISEAR dataset. Performance of BERT model (98.72%.) surpasses the accuracy achieved by other algorithms examined in this research and results reported in the existing literature. Importance, applications and obstacles in the area of emotion recognition are also briefly discussed in this paper. Evolution of emotion is surveyed and discussed along with emotion models. The paper also delves into the available datasets for emotion analysis.

2015 International Conference on Technologies for Sustainable Development (ICTSD), 2015

Proceedings of 3rd International Conference on Reliability, Infocom Technologies and Optimization, 2014

2015 International Conference on Technologies for Sustainable Development (ICTSD), 2015

International Journal of Applied Engineering & Technology (London), 2023

Emotion gives thoughts and feelings of a person. Emotions are reflected from text, speech, gestur... more Emotion gives thoughts and feelings of a person. Emotions are reflected from text, speech, gestures as well facial expressions. Recognizing and interpretation of emotion from such reflections plays an essential task in the interaction between human and machine conversation. Different views or opinions in terms of sentiments are derived from such emotions. Sentiment analysis provides negative, positive, or neutral terms; whereas emotional analysis provides deeper analysis of participant's emotions that tries to drill down into the various user behaviors. Emotion Recognition and Sentiment Analysis together can help to identify, process and illuminate human affects. In this study, a diverse array of algorithms, including LSTM, BiLSTM, CNN, and RCNN, as well as ML algorithms such as RF, MNB, SVM, and Logistic Regression, were employed for the purpose of recognizing emotions from text. Models were trained and tested along with different word embedding methods and activation functions. Well balanced ISEAR dataset (International Survey On Emotion Antecedents and Reactions) is utilized in experimentation. Implementation of a fine-tuning and evaluation process for a BERT-based model using the Hugging Face Transformers library is also done for ISEAR dataset. Performance of BERT model (98.72%.) surpasses the accuracy achieved by other algorithms examined in this research and results reported in the existing literature. Importance, applications and obstacles in the area of emotion recognition are also briefly discussed in this paper. Evolution of emotion is surveyed and discussed along with emotion models. The paper also delves into the available datasets for emotion analysis.

2015 International Conference on Technologies for Sustainable Development (ICTSD), 2015

Proceedings of 3rd International Conference on Reliability, Infocom Technologies and Optimization, 2014

2015 International Conference on Technologies for Sustainable Development (ICTSD), 2015