sentiment analysis, natural language processing, speech recognition, information retrieval, machine learning, user experience.
The popularization web-based multimedia content has raised the need to analyze and retrieve it automatically. There is an immense need for identification as well as classification of sentiments (human emotions) due to unfeasibility of labelling big data manually on a larger scale. The Proposed system is to automate the content identification using Audio Processing and Machine Learning. The system aims to extract the audio stream from any multimedia as the input. Using natural language processing, the system will automatically generate raw text useful for identification of the type of multimedia. Further using this data and machine learning, the system will label the content with the opinion of the speaker as output. It will show various sentiments along with their intensity.