Tag it:
Delicious
Furl it!
Spurl
NewsVine
Reddit
YahooMyWeb
Technorati

ChordMiner

screenshot5

The idea of ChordMiner is to analyze guitar chord sequences on the web in order to support guitar players in finding adequate chords. For a given chord, ChordMiner shows the chords that most likely follow this chord, the most interesting chords and typical chord sequences in which the chord appears. ChordMiner also provides a filtering option that allows to prune potentially long lists of relevant chords by substring matching. Finally, ChordMiner provides a history that enables musicians to interactively create chord sequences, e.g. while composing.

This application is meant as a demonstration of the power of text mining. ChordMiner is able to produce relevant recommendations, almost without any knowledge of musical theory. Furthermore, these recommendations do not only reflect relationships among chords, but also guitar playing practice. Chords that are easy to play will usually be recommended more often than such that are harder to play. Again, there is no background knowledge to guide this ranking. Everything is achieved by the skills of the people that composed and transcribe the music, and by some simple statistical methods.

Contact

ChordMiner was developed by Michael Wurst on the basis of the open source data mining tools Rapid-I and WVTool. For more information on this project, suggestions or comments, please contact us.

 
Valid XHTML 1.0 Strict Valid CSS!