DocumentCode :
3498734
Title :
Solely Tag-Based Music Genre Classification
Author :
Zhen, Chao ; Xu, Jieping
Author_Institution :
Comput. Sci. Dept., Renmin Univ. of China, Beijing, China
Volume :
1
fYear :
2010
fDate :
23-24 Oct. 2010
Firstpage :
20
Lastpage :
24
Abstract :
As a fundamental and critical component of music information retrieval (MIR) systems, automatically classifying music by genre is a challenging problem. The approaches depending on low-level audio features may not be able to obtain satisfactory results. In recent years, the social tags have emerged as an important way to provide information about resources on the Web. In this paper we are interested in another aspect, namely how perform automatic music genre classification solely depending on the available tag data. Two classification methods based on the social tags (including music-tag and artist-tag) which crawled from Last. fm are developed in our work. The first one, we use the generative probabilistic model Latent Dirichlet Allocation (LDA) to analyze the music-tag. Then, we can compute the probability of every tag belonging to each music genre. The starting point of the second method is that music´s artist is often associated with music genres more closely. Therefore, we can calculate the similarity between the artist-tags to infer which genre the music belongs to. At last, our experimental results demonstrate the benefit of using tags for accurate music genre classification.
Keywords :
Internet; classification; information retrieval; information retrieval systems; music; probability; Web; latent Dirichlet allocation; music information retrieval systems; social tags; solely tag-based music genre classification; LDA; artist tag; music genre classification; music tag;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Web Information Systems and Mining (WISM), 2010 International Conference on
Conference_Location :
Sanya
Print_ISBN :
978-1-4244-8438-6
Type :
conf
DOI :
10.1109/WISM.2010.152
Filename :
5662276
Link To Document :
بازگشت