Output details
13 - Electrical and Electronic Engineering, Metallurgy and Materials
Queen Mary University of London : A - Electrical and Electronic engineering
Music Information Retrieval Using Social Tags and Audio
Paper is one of the earliest marrying text-based with content-based Information Retrieval for musical audio. It demonstrates that where tracks have little or no textual metadata (are not popular, from little-known artists, long tail), they can still be reliably retrieved. Paper was early outcome of OMRAS2 one of EPSRC's first ICT Large Grants (EP/E017614/1) that produced over 120 papers (omras2.org), collaborative between Queen Mary and Goldsmiths. It led to Levy's employment with last.fm, the world leader in social tagging of music, to 3 research internships with the company and subsequent research on automatic ontology generation (DOI: 10.1109/TASL.2013.2263801).