以個人化距離公式為基礎之音樂推薦系統─使用基因規劃法

在一般的音樂推薦系統中，為了對使用者進行推薦，必須有辦法針對不同的音樂，進行相似度評估，藉以求得兩兩音樂間的距離，以作為推薦時的依據。對於如何計算音樂間的距離，已有許多學者提出了不同的策略與方法。本研究假設在使用者的心目中，對於不同音樂間的相似感受度，隱含了一套專屬於個人的距離評估公式。為求得此公式，系統將先針對使用者在不同歌曲間的相似度感受度進行詢問，接著利用基因規劃法求取使用者之個人化距離評估公式。所求得的公式搭配自行設計的清單生成機制，將可產生新的推薦清單供使用者使用。透過實驗，將本研究所提出的公式與常見的歐幾里德距離公式進行比較（含權重式），結果證明，本研究所提出的距離公式，對於相似音樂的推薦將更為精確。

關鍵字

音樂推薦；距離；基因規劃

並列摘要

In general music recommendation systems, there must be a way to measure the similarity between different pieces of music and then obtain the distances between them. The distances will then be the basis for recommendation. Many scholars have proposed different strategies and methods to calculate the distances between pieces of music. This study assumes that, in a user’s mind, there is a set of personalized distance formula for measuring the similarity that he or she feels between different pieces of music. To obtain this formula, the system first inquires about the similarity that the users feel between different songs, and then runs genetic programming method to acquire the users’ personalized distance formulas. New recommendation lists for users are subsequently generated from both the obtained formulas and the self-designed list generation mechanism. Through experiments, we compared the formulas proposed in this study with the common (weighted) Euclidean distance formula. The results showed that the distance formulas proposed in this study were more precise for similar music recommendation.

並列關鍵字

Music Recommendation ； Distance ； Genetic Programming

參考文獻

[1] 蘇軍維 (2009)：基於權重式歐幾里德距離之音樂推薦系統設計。國立屏東科技大學資訊管理系碩士論文。

[24] Ragno, R., Burges, C. J. C. & Herley, C. (2005). Inferring Similarity between Music Objects with Application to Playlist Generation. In Proceedings of the 7th ACM SIGMM International Workshop on Multimedia Information Retrieval (pp. 73–80).

[3] Cohen, W. W. & Fan, W. (2000). Web-Collaborative Filtering: Recommending Music by Crawling the Web. Computer Networks, 33(1-6), 685–698.

[5] Goldberg, D., Nichols, D., Oki, B. M. & Terry, D. (1992). Using Collaborative Filtering to Weave an Information Tapestry. Communications of the ACM.

[6] Hayes, C. & Cunningham, P. (2000). Smart Radio-Building Music Radio on the Fly. Proceeding of Expert Systems 2000.

國際替代計量

以個人化距離公式為基礎之音樂推薦系統─使用基因規劃法

全文下載

主題瀏覽