Salsa dataset: first salsa music knowledge base

A well-known musical genre in our region and that is part of our cultural identity Caleña is salsa. Its origin dates back to the 30s of the last century and since then this genre has been modified by different artists in different regions of the world, each with a different vision of it, with cultural experiences and with the contribution of new instruments and new technology. This makes salsa an intrinsically complex genre and difficult to define in qualitative terms. Despite the growing popularity of the genre in the world, salsa has not been analyzed from the computational point of view to derive what are the acoustic components that define and differentiate it from other musical genres. In this document we will discuss the results of the process of creating a knowledge base of salsa music that is freely available to the scientific community and that collects the acoustic information of more than 20,000 songs of that musical genre. With this information, which characterizes the acoustic signal, it is intended to model different characteristics of the genre through artificial intelligence techniques. In addition to making this information accessible to researchers interested in salsa music, another important contribution of this project is to provide tools for the knowledge base to grow with the help of the scientific community. For this, a software was developed that extracts the pertinent acoustic information of songs that users have to be sent and added to the knowledge base.

• Autores: Sarria M., Gerardo M.; Mora, Mario Julián; Arce-Lopera, Carlos
• Fecha de publicación: 2016
• Institución: Universidad EAFIT
• Repositorio: Repositorio EAFIT
• Idioma: spa
• OAI Identifier: oai:repository.eafit.edu.co:10784/15264
• Acceso en línea: http://hdl.handle.net/10784/15264
• Palabra clave: Musical Genre Salsa; Dataset; Music Information Recovery; Género Musical Salsa; Dataset; Recuperación De Información Musical
• License: Copyright (c) 2016 Gerardo M. Sarria M., Mario Julián Mora, Carlos Arce-Lopera