Audio and Text Sentiment Analysis of Radio Broadcasts

• Published in: IEEE access Vol. 11; p. 1
• Main Authors: Dhariwal, Naman, Akunuri, Sri Chander, Shivama, Sharmila Banu, K
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01-01-2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Analytical models; audio; Audio data; Data mining; Data models; Emotion recognition; Emotions; Natural language processing; news; Public opinion; Radio broadcasting; radio broadcasts; Sentiment analysis; Social networking (online); Speech recognition
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2023.3331226

The rapid growth of radio broadcast services has created a vast amount of audio data that can provide insights into public opinion and emotions. This research extends the boundaries of sentiment analysis to audio data and aims to propose a computational approach which is termed as bifurcate and mix, for sentiment analysis and emotion detection that leverages advanced natural language processing techniques from audio sentiment analysis tools like Vokaturi, transcription services like AssemblyAI and text sentiment analysis lexicons like VADER to extract and categorize sentiments from the audio data to generate a more efficient real-time sentiment analysis model. The results of the analysis reveal patterns and trends in the sentiments expressed in radio broadcasts by the News Service Division: All India Radio - 'Akashwani'. This research's methodology will contribute to the development of novel applications for sentiment analysis in the media industry and provide valuable insights into public opinion and emotions.