Computational analysis workshop: Comparing four approaches to melodic analysis

Abstract

We compare four computational approaches of melodic analysis according to diverse approach aspects: input type (monophonic or polyphonic), pattern identification type (strict or similar), analysis segmentation, aim of approach, motivic pattern representation, and type of result representations. The considered four computational approaches are the following: a similarity neighbourhood approach by Adiloglu (Adiloglu and Obermayer 2006a, b), a multiple viewpoint representation and discovery approach by Anagnostopoulou (Anagnostopoulou, Share and Conklin 2006), a topological approach by Buteau (2005), and an approach based on multidimensional closed pattern mining by Lartillot (Lartillot and Toiviainen 2007). © 2009 Springer-Verlag.