Computing with words Concept Applied to Musical Information Retrieval[ELSEVIER]英文文献.pdfVIP

URL: http://www.elsevier .nl /lo ca te/ ent cs/ vol ume 82. htm l 12 pages Computing with words Concept Applied to Musical Information Retrieval Bozena_ Kostek 1 Sound and Vision Engineering Department Gdansk University of Technology Narutowicza 11/12, 80-952 Gdansk, Poland Abstract The ob jective of the pap er is to provide cognitive-based mechanisms underlying pro cessing of musical instrument sounds. The system prop osed by the author based on the rough set metho d and on fuzzy logic provides knowledge on how humans internally represent such notions as quality and timbre and therefore it allows for the human-like automatic pro cessing of musical data. Therefore Computing with words concept can b e used in musical information retrieval domain by oering b et- ter pro cessing of sub jective descriptors of musical instrument sounds and enabling the analysis of data that would result in extraction of semantic information related to musical instrument sounds. This pap er shows rst a review of developments in the domain of timbre mapping and classication. A decision table is built of se- mantic descriptors of musical instrument sounds, then rules extracted by the rough set metho d and the pro cessing of musical timbre based on fuzzy logic is shown. An example of rough-fuzzy pro cessing is given and conclusions are derived. Key words: Computing with words, soft computing, rough sets, fuzzy logic, Musical Information Retrieval, MPEG7. 1 Intro duction One of p ossible examples of Computing with words pro cessing is the as- sessment of the quality of musical instrument sounds, for which humans use criteria that are rarely quantitative but most often qualitative. T