Ranqueamento de Acordes de Violão com Base na Dificuldade: Um Conjunto de Dados de Tríades com Pontuações de Dificuldade

Abstract

A six-string guitar offers vast possibilities for chord formation due to its scale composition, reflecting the instrument’s flexibility and adaptability. However, some chord shapes can be challenging, influenced by factors such as f inger placement, hand stretch, the presence of barre chords, neck position, or the guitarist’s familiarity with the chord. In this work, we introduce a dataset consisting of 500 chord shapes in standard tuning, specifically major and minor triads, ranked by their execution difficulty. The initial ranking was determined using objective criteria established in the literature, considering the comple xity of chord execution, and was refined with pairwise comparisons made by ten musicians of varying proficiency levels, using the ELO ranking system. This sys tem allowed for dynamic adjustment of rankings based on musician feedback, ensuring a refined difficulty level. The comparison process was optimized to only compare chords with close difficulty levels, reducing the number of pairs to be evaluated. Spearman’s correlation analysis confirmed a high agreement between the initial criteria and human perceptions, with a value of 0.99001 and a p-value of 0. The final dataset is a valuable resource for the Music Information Retrieval community and the creation of personalized recommendation systems in music learning, which can adapt content to user experience, facilitating pro gressive learning for guitarists through educational tools that often overlook chord difficulty