Informatics, TU Vienna

Web Content Mining for Music Retrieval

Automatically extracting information relevant to a specific topic or domain from the Web is a hot research topic, not only in traditional text-based information retrieval, but also for multimedia retrieval tasks.

Abstract

Automatically extracting information relevant to a specific topic or domain from the Web is a hot research topic, not only in traditional text-based information retrieval, but also for multimedia retrieval tasks. In this talk, I will focus on the subdomain of music retrieval and will present different Web-based information extraction techniques aimed at enriching traditional music applications, but also to build the foundation for innovative, next-generation music services, such as personalized music recommenders. In particular, methods that address the following research problems will be presented:

  • artist similarity measurement using Web pages and using microblogs (tweets)
  • artist popularity estimation using cultural features from Web-based sources
  • intelligent user interfaces to browse music collections via acoustic clustering and Web-based feature extraction

For the first topic, I will report on large-scale evaluation experiments aimed at determining well-performing parameter settings for modeling the Web-based music similarity space (e.g., TF-, IDF-formulations, normalization strategies, similarity functions). As for estimating an artist's popularity, different techniques that rely on various data sources (Web, Twitter, last.fm, Peer-to-Peer Networks) will be presented and discussed. Finally, a prototype implementation of an intelligent user interface to browse music collections will be presented. The "nepTune" interface combines content-based features to cluster music according to acoustic properties and contextual features mined from the Web to facilitate orientation and browsing within the virtual music landscape the interface is based upon.

Biography

Markus Schedl graduated in Computer Science from the Vienna University of Technology. He earned his Ph.D. in Computational Perception from the Johannes Kepler University Linz, where he works at the Department of Computational Perception. His main research interests include Web Mining, Music and Multimedia Information Retrieval, Information Visualization, and Recommendation/Personalization. He is also co-founder of the International Workshop on Advances in Music Information Research.

Note

This talk is organised by the Knowledge Based Systems Group at the Institute of Information Systems, supported by the Center for Computer Science and the Wolfgang Pauli Institut (WPI).