Sentiment Prediction Using Voice

Sentiment Prediction Using Voice

Jaya Vishwakarma¹, Keshav Sharma²

1,2 Students, Department of Computer Science and Engineering, Babu Banarasi Das Northern India Institute of Technology

Dr. Anurag Shrivastava

Head of Department of Computer Science & Engineering, Babu Banarasi Das Northern India Institute of Technology

---------------------------------------------------------------------***-------------------------------------------------------------------

Abstract - The project ‘Sentiment Analysis using Voice’ will help in identifying the fraud calls like scammers asking for bank details, personal details of less educated people ,who believe these frauds easily and believe they are actually the officials from the agency ,and got scammed by them .The limitations to this can be noise and disturbance in audio data and background noises to minimize the impact of these errors we are working on a model to provide a viable solution. The project can also be used in detecting negative emotions and harmful thoughts on more research and on including some additional features. The project will come in handy when used on calls for real time analysis of call, for detecting the call is fraud or valid on time. It will start from basic after learning process , understanding sound and it’s features in machine from different sources , and how to handle sound data , how to preprocess it and how to build model for it. The final step involves deployment and testing of model on real voice data provided by one of our team members.

Key Words:  sentiment, voice, light, dataset, machine learning, testing.