Knowledge discovery in musical databases for moods detection

Abstract

In this paper, methodology Knowledge discovery in databases is used in the design and implementation of a tool for moods detection from musical data. The application allows users to interact with a music player, and based on their playlist and musical genre, recognizes and classified their emotional state using a neural network. The results found are promising to have an accuracy of more than 72,4%, in addition the developed tool allows the constant taking and storage of data, the analysis in real time and issues suggestions of songs to positively influence the current emotional state, so that a greater use of the application can guarantee better results.