David Rizo

Interactive Music Analysis

Automatized computer music analysis has been performed so far mainly using classical artificial intelligence / pattern matching techniques with little or no user interaction in the decisions. Many of the problems that lie inside the music analysis area either don't have a unique valid solution, or this solution can be classified more as an exception than a rule. In recent years, the theoretical studies on interactive pattern matching have proved that for this sort of situation, the interaction of the user can lead to new levels of accuracy besides increasing the usability of the systems. In this work we explore the use of interactive pattern matching techniques to two problems: tonal music analysis focusing on the melodic analysis phase, and the similarity analysis between works.