• Neseebha
• Jithendra P.K Nayak

Among the most crucial areas of research in human-computer interaction is sentiment analysis using multimodal data, including text and speech. The goal of this endeavour is to create an effective sentiment recognition system utilising methods from machine learning to categorise emotions from text and spoken audio input, especially happy and neutrality. The system can efficiently analyse and forecast emotional states by fusing real-time speech-to-text conversion with Natural Language Processing (NLP) methods including TF-IDF and Naive Bayes classifiers. Entails sing a publicly accessible emotion-labeled dataset for extraction of characteristics, model training, data preprocessing, and evaluation. Experimental results demonstrate satisfactory accuracy, highlighting the potential of integrating speech and text modalities for improved emotion recognition. This system can be used in various applications include mental health monitoring, smart assistants, and human-robot interaction.

Sentiment analysis, Emotion recognition, Multimodal data, Speech emotion recognition, Text emotion recognition, Speech-to-text conversion, Natural Language Processing (NLP), Machine learning, Naive Bayes classifier, TF-IDF, Feature extraction, Model training, Data preprocessing, Happy and neutral emotion classification.

"SENTIMENT ANALYSIS NEUTRAL AND HAPPY USING DEEP LEARNING", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.971-976, July-2025, Available :http://www.jetir.org/papers/JETIRGX06180.pdf

"SENTIMENT ANALYSIS NEUTRAL AND HAPPY USING DEEP LEARNING", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. pp971-976, July-2025, Available at : http://www.jetir.org/papers/JETIRGX06180.pdf