Mood Identification Using Machine Learning

We can search a song by different tags such as artist, name of song, album, band, genre etc. but this type of search is restricted when it comes to search a song according to the mood of a person at that instant. So, a new parameter had to be added to so that the person can choose a song according to his mood. But it is not that simple to implement this idea so we have to interpret the music type with this mood tag.Most of the time this tag is manually edited which is prone to any errors. So we want to build a system that automatically detects the mood for a particular song and for a particular instant.. It will lower the errors and manual work will be avoided. Music can be organized in a better way according to the user's mood, with this way. Many different sources came into the picture after researches, with different solutions to the same problem that is identifying mood from song's text file which provides effective data which is different from that is fetched from music piece. So, we will make a hybrid model that combines both audio features and lyrics to find out the mood of a song effectively. Moreover we are working on Bollywood music which has many flaws.