PDF Publication Title:
Text from PDF Page: 007
G Model POETIC 1239 No. of Pages 13 (]GIF$DT)1_.giF[6 M. Airoldi et al. / Poetics xxx (2015) xxx–xxx Fig. 1. YouTube related-music videos network. Colours indicate different clusters. We ran a community detection algorithm (see Van Meeteren et al., 2010) in order to identify clusters of videos internally connected with each other within the overall network. This procedure results in a ‘modularity’ score, which can be interpreted as the indicator for the ‘goodness of fit’ of this decomposition (Newman, 2006). We note this is quite high (0.792) and therefore interpret this as a legitimation of the results of the community detection (Gaul & Klages, 2013). The number of detected clusters is 50, and these clusters are marked using the different colours in Fig. 1. Inspection of this visualisation suggests a fairly high level of aggregation, considering that the dataset is composed of more than 20,000 videos. Furthermore, the size of the majority of clusters is approximately of the same magnitude. This provides additional legitimation to our overall interpretation. After having detected 50 groups of music videos that have been co-viewed by a vast number of YouTube users, we moved to the second phase of our data analysis in which we investigated the presence of common traits shared by music videos belonging to the same cluster. In order to do so, we first analysed the ‘semantic core’ of each cluster that is, the ten most frequent words in the video titles, excluding irrelevant ‘stop-words’ (see Krippendorff, 2013). We did this by labelling each cluster on the basis of the main commonalities shared by its videos, which inductively emerged from the textual content of the titles. Note that this normally summarises the content of the video with a few keywords in order for it to be easily searchable through simple queries. For instance; the ten most recurrent words in Cluster 0 (2049 videos overall) are ‘cover’; ‘One Direction’; ‘live’; ‘lyrics’; ‘Daft Punk’; ‘Ariana Grande’; ‘Katy Perry’; ‘Justin Bieber’; ‘Pharrel Williams’ and ‘Miley Cyrus’. Overall; 48% of the videos in Cluster 0 feature at least one of these words in the title. Among these videos; 170 are cover versions of contemporary international pop songs; while the large majority is constituted by official videos of commonly recognised ‘mainstream’ pop artists. The most viewed content of this cluster is Miley Cyrus’ ‘Wrecking Ball’; while the most connected (that is; the one featuring the highest number of ties; particularly central in the cluster’s graph) is ‘Gorilla’ by Bruno Mars. Thus; we labelled this cluster ‘Teen Pop’ not because all its music videos can be considered exactly so; but simply because this definition fits most of its content (e.g. One Direction; Ariana Grande; Justin Bieber). The genre labels employed to name the clusters are either derived directly from video titles or; especially in the case of clusters where artist names are more common than genre names; are attached by the authors according to conventional music classifications. Nevertheless; the primary purpose of using such labels is to provide the reader with a rough idea of what these clusters are made up of. They therefore should not be taken as comprehensive categorisations (Table 1; below). Clusters are internally homogeneous from a semantic point of view. In 27 clusters out of 50, the ‘semantic core’ (i.e. the ten most recurrent relevant words in the titles) is shared by the totality of videos. Also, in the specific case of ‘Sounds of Nature’ the internal semantic relatedness is particularly strong, since 4 out of 10 of the most recurrent words are present in Please cite this article in press as: M. Airoldi, et al., Follow the algorithm: An exploratory investigation of music on YouTube, Poetics (2016), http://dx.doi.org/10.1016/j.poetic.2016.05.001PDF Image | Follow the algorithm: An exploratory investigation of music on YouTube
PDF Search Title:
Follow the algorithm: An exploratory investigation of music on YouTubeOriginal File Name Searched:
AiroldiBeraldoGandini2016Preprint.pdfDIY PDF Search: Google It | Yahoo | Bing
Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info
Cruising Review Topics and Articles More Info
Software based on Filemaker for the travel industry More Info
The Burgenstock Resort: Reviews on CruisingReview website... More Info
Resort Reviews: World Class resorts... More Info
The Riffelalp Resort: Reviews on CruisingReview website... More Info
CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com (Standard Web Page)