BandNavi: Artists discovery system using member change information

We describe the technique for automatically constructing the network which indicates the relationship between bands and musicians from the Web pages. And we also present the applications of the network for the recommendation and the classification of bands. The relation between bands and musicians make complex network structures according to the repetition of member change and appearance of the guest musician. We constructed the network by automatic acquisition of member name from Web pages collected by using Google Search API. In addition, we defined strength of the relations between bands from the number of duplication of members by using the constructed network. And we attempted to construct  systems of the recommendation and  the hierarchical clustering of bands.

A band-member backtracking application called BandNavi, BandNavi HD are available on App Store, which enables a user to discover new songs and bands by tracing musicians who have played in different bands.

バンドネットワーク
▲Band Network

 

bandnet_2

▲A network on a time line

 


fig4
▲Searching system for bands

group
▲Grouping of bands

References

  • 吉谷幹人, 宇佐美敦志, 浜中雅俊: “BandNavi: バンドメンバーの変遷情報を辿るアーティスト発見システム,” 第18回インタラクティブシステムとソフトウェアに関するワークショップ, pp. 29-34, July, 2010. [PDF]
  • 吉谷幹人, 宇佐美敦志, 浜中雅俊: “BandNavi: バンドメンバーの変遷情報を辿るアーティスト発見システム,” 情報処理学会 音楽情報科学研究会 研究報告 MUS-86-16 July, 2010. [PDF]
  • 吉谷幹人, 宇佐美敦志, 浜中雅俊, “メンバー情報に基づくバンドネットワークの構築と利用”, 情報処理学会 音楽情報科学研究会 研究報告 2009-MUS-82-5 (2009). [PDF]

※ Copyright
IPSJ holds copyright of papers above. Please follow the guidelines HERE.