Music Genres

From dandi08

Jump to: navigation, search

Contents

[edit] Overview

An ontology of music genres would allow for an in depth look at music genres that went beyond the standard relationships that exist within most commercial music programs. It was suggested by a student in the class that in their experience with Apple's iTunes they had trouble with the genre parameter. Music could only be placed in single genre categories. For instance, the 19th century composer Cluade Debussy was considered to be a member of the Romantic Period of Orchestral music. While a student of such music would consider Debussy's music to be drastically different than the compositions of Bach, a Baroque composer, a casual listener would simply categorize this music as "Classical." However, both the student and the casual listener would be able to tell the difference between these composers and The Beatles. Furthermore, while the casual listener would classify both of the Orchestral composers as "Classical," the music student would say that neither were "Classical" composers as neither were composing music during the "Classical" Period, the period that is best known for Beethoven's Symphonies. An Ontology of these sorts would correct these discrepancies.

[edit] Project Points

The main areas that this project would cover are:

  1. A taxonomy of genres, as in "Swing" is a sub genre of "Jazz" In order to avoid disambiguation, genres would be taken, as often as possible, from a credible source such as iTunes. The current iTunes Store setup includes a genre, sub genre system, so lots of ideas are already out there
  2. Instances of music groups (Artists, Composers, DJ's etc...) that fit into these genre categories. Again, in order to avoid disambiguation the genre of the artist should be determined by a credible source, or preferably, how the artist defines themselves and matched to the best example within the Ontology.
  3. Relationships between classes other than "genre X is a sub genre of genre y." For instance, Rock music can be considered to be derived from blues, however they are drastically different
  4. Properties of the particular music groups and classes.

[edit] Genre Taxonomy

It will be important to identify the main properties of the particular genres. This could include the period that music was mainly written in that style (it could be open ended as some music is still actively played) as well as how that music evolved from a previous genre. It will also be important to determine how specific genres should be. Surely it would be important to include "Alternative Rock" however a term like "Post Slater-Kinney Olympia Rock" might not be a very useful category. A tentative breakdown might be something that looks like Rock>90's Alternative> Northwest Rock. This might be a poor example as Northwest Rock exists outside of just the 90's.

[edit] Instances

It will be important of course to provide examples of what music groups fit into what categories. However this should not be the main focus of the project as in the earlier stages of the project it would be better to have a shell that, once completed, would allow for easy input of artists into the ontology. It will also be important to avoid a situation of an artist being "a better example" of a group than other artist. No fuzzy set theory.

[edit] Relationships

A general list of relationships is as follows:

  1. Classes: This would include Genre, Sub-genre and so on. Classes at a higher level would be more music promoter oriented (ie. the big 4) while genres at lower levels would be more artist driven.
  2. Evolved from: Could as be referred to as influenced, this would include relationships like "Rock music came out of blues music" This would allow the ontology to be placed in a chronological order, which would allow for analysis of trends in music.
  3. Audience: More of an instance-specific relationship, this would include who the music was directed towards. For example the "Michael Jackson" instance is a member of Motown however it can be considered "pop" music as it was directed towards a more mainstream audience. This feature would allow a look at how music trends have existed in the past.