Predictive Visual Analysis of Speech Data using Machine Learning Algorithms

Abstract

Audio sense analysis using automated speech recognition is an emerging research field where opinions or feelings are natural audio displayed by a speaker is detected. Still, text- based sentiment detection is relatively unexplored and very less used in physical word despite its huge possible implementation area because Talking and listening improves information if it is nice then it encourage to listen to you which helps better understanding the things and to form a relationship. Talking have many emotions in it; this is not just nice thing or good news, but also anger, fear, criticism, or blame hence can be categories in positive, neutral or negative communication. In this new architecture, we perform a predictive visual analysis of big data collecting from various speaking source and here we perform graphical and non-graph analysis and visualize it using machine learning algorithms (Naive Bayes). And we perform weighted word cloud visualizations, which give improved semantic insights. The results obtained by this work can help in finding good teacher, orator, speaker, and examine self-talking skill.