raghuvenmarathoor / MusicSuggestor

Song suggester is an innovative application which will learn the behaviour of the listener while playing songs. The application will use finger printing methodology of the Echo Nest (An open source database having records of over 30 million songs) to identify songs in the application. In this Project there will be a music player with various options which will play mp3 files from hard disk of the client or user which helps to refer sources from the directory of the system and to play the song through the system’s output. The system will analyse the sample of the file it is playing and produces the code string using the code generation tool made available by echo nest. This code string will be sent to the web server where the details of all the identified music will be saved. The web server is maintained to collect all the music identified by the users of this application, which will have a user account in this web server. This web server will have the API which is required for the precise identification of the song from Echonest. Beyond that, the web server will analyse the behaviour of the user’s listening habit and to suggest songs accordingly. The Song Suggestion part will have algorithms for learning the user’s listening habits through identification of songs and the suggestions of songs according to the user’s selection of classifications like most popular, user specific favourites etc. The website will use the web server’s database to analyse a user’s taste and help them to create suggestions for groups, friends and discussions. It will also help the user to analyse the top hits and new trends in music. The proposed system will help the user to get the most appropriate song he likes, information about new releases and artists. The user can search for suggested songs in the internet; it recommends songs according to the criteria of the user and provides opportunities to learn the user’s listening habits.