This paper focuses on the development and implementation of pattern matching algorithms designed for the analysis of musical sequences. The primary objective is to create algorithms capable of efficiently and accurately identifying instances of musical patterns within a dataset encompassing both simple compositions and well-known musical pieces. This goal is achieved through the adaptation and extension of the Tuned-Boyer-Moore algorithm, coupled with the introduction of δ- and δ, γ-approximation techniques. The performance of these algorithms is evaluated on a dataset containing both self-generated pieces as well as well-known pieces. The algorithm consistently was able to detect both exact and approximate pattern occurrences accurately, even when the pieces were subject to changes in rhythm and key. A series of testing rounds involving manipulation of δ and γ values, showcases the algorithms’ adaptability and efficiency. |
*** Title, author list and abstract as seen in the Camera-Ready version of the paper that was provided to Conference Committee. Small changes that may have occurred during processing by Springer may not appear in this window.