Thesis Proposal
Title: A Mood-Based Music Classification and Exploration System
Abstract:
Mood classification is very subjective. There have been many psychological studies performed throughout the past century relating to music and emotion, and as a result there exist many different representations of human emotion. This thesis proposes to utilize the best possible psychological model of emotion by incorporating the findings of these studies into an innovative front-end for a digital music library where music can be queried, browsed, and explored by mood, rather than artist, album, or genre. With the use of state-of-the-art audio and textual analysis tools this work proposes to automatically and accurately classify music by mood.
+++++++++++++++++++++++++++++++++++++++
Let me know what you think of the idea!
Abstract:
Mood classification is very subjective. There have been many psychological studies performed throughout the past century relating to music and emotion, and as a result there exist many different representations of human emotion. This thesis proposes to utilize the best possible psychological model of emotion by incorporating the findings of these studies into an innovative front-end for a digital music library where music can be queried, browsed, and explored by mood, rather than artist, album, or genre. With the use of state-of-the-art audio and textual analysis tools this work proposes to automatically and accurately classify music by mood.
+++++++++++++++++++++++++++++++++++++++
Let me know what you think of the idea!
This entry was posted by meyers, on
Saturday, January 27, 2007.
You can leave your response.
owen - looks like an interesting topic (and one that I am quite interested in) - I have some thoughts on collecting mood ground truth data - perhaps we should talk - give me an email if you'd like Paul.Lamere@sun.com.
leave a response