The Fractal Dimension of Music: Geography, Popularity and Sentiment Analysis

Explaining Successful Docker Images Using Pattern Mining Analysis

“…Using this network delineates four typical social behavior patterns. Further domains where the analysis and prediction of success is a challenging task are music [19], movies [11] and school performances [7]. However, to the best of our knowledge, our approach is the first that is based on complex descriptions such as those of Docker images, and which tries to understand the reasons of popularity and endorsement.…”

The italian music superdiversity

“…In [26] a lyrics dataset based on Valence-Arousal model of Russell [35] is created employing (AMG) tags. Likewise [13,31] and the work we present, Affective Norms for English Words (ANEW) [9] is used as a lexical resource. Once classified AMG tags in the four Russell's model quadrants using ANEW, songs are categorized using the obtained tags, and finally, annotations are evaluated by employing human evaluators.…”

“…Indeed, we extract each medoid song as the most artist's representative song identified by minimizing the sum of the Euclidean distances between the artist's tracks' musical features gathered from Spotify. Given the results obtained in [30] and [31], the medoids computation doesn't take into account all the ten musical features obtained from Spotify. Indeed, previous results display that features like speechiness, liveness, loudness, and tempo present similar values in each type of data aggregation and comparison, while others features are conversely high discriminant.…”

“…Music's origins, like those of language, are hidden in the most ancient past of the humanity's history. For centuries music has been part of human civilization and each culture has given birth to its own music [31]. However, nowadays, the globalization and the progressive reduction of geographical distances facilitated by ubiquitous of the World Wide Web and the media has brought the collapse of the music barriers.…”