Solving the long tail phenomenon of amplayer

• Main Author: Perez, Royce Joseph G.
• Format: text
• Language: English
• Published: Animo Repository 2015
• Subjects: Music; Computer Sciences
• Online Access: https://animorepository.dlsu.edu.ph/etd_bachelors/2965
• Institution: De La Salle University

AMPlayer is a portable music recommendation system that plays certain music that regulates the arousal level of the child in relation to the subjects electrodermal activity (EDA) and the songs music features. This research focuses on creating a Music Information Retrieval System on the AMPlayer that recommends songs to the playlist by downloading from a music online database. It expands its data to improve its music recommendation automatically. The Music Information Retrieval System uses AMPlayers user model. From there, the music that will be recommended will extract these music features namely spectral flux, spectral centroid and its variability, zero-crossings, root-mean-square, Mel-Frequency Cepstral Coefficients (MFCC), and beat sum. These features are chosen because it impacts the emotions, i.e. express emotions, and create influence on musical moods. This study also validated the impact of music on individuals with autism, which has been confirmed as per the results of the investigation. The investigation proves that different music genre affect individual with autism, investigating and extracting features of different music, and looking at the subjects EDA it shows that different kind of music yields different kind of electrodermal activity of a individual with autism. Its specific effects and music features that induce arousal, however, remain unknown. In addition, creating a new module for the Music Information Retrieval System did improve the recommendations of music and solve the long tail phenomenon problem from the previous system by having a large music database for recommendation that can enhance the overall use and experience of the system. The usability of the new module has confirmed as above average as per the results of the survey. Ten usability test questions were delivered for every participant that will answer the survey form and every question in the survey form resulted above 60% in rating the system usability. The survey that has been used is created by John Brooke in 1986 for measuring the system usability scale of a system. System Usability Scale (SUS) survey has since been tested on hardware, consumer software, websites, cell-phones, IVRs and even yellow pages. This survey is participated in by parents, teachers and caregivers.