Emotion Detection in Punjabi Audio: A CNN-Based Sentimental Analysis

Le résumé fourni par la source

Sentiment analysis, often called opinion mining, stands as a critical subfield within Natural Language Processing (NLP), specializing in the interpretation of emotions and viewpoints in various communication forms. Though sentiment analysis has traditionally excelled in text data, it’s now extending its reach to audio data. This conference article offers a thorough investigation into the application of sentiment analysis in the context of Punjabi audio content, an underrepresented regional language in sentiment analysis research. Employing Convolutional Neural Network (CNN) architecture, we conducted an experimental study involving audio samples from 50 participants, spanning a spectrum of emotions, including positive, negative, and neutral. The experimental results showcased a notable accuracy rate of 92%, underscoring the adaptability and robustness of sentiment analysis, particularly in regional languages. This research not only broadens the scope of sentiment analysis into audio data but also emphasizes the significance of understanding sentiment in less-represented languages, with practical implications across diverse domains, from marketing and social media monitoring to customer feedback analysis.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.