Understanding the dynamics of on-line content popularity is an active research field with application in sectors as diverse as media advertising, content replication and caching and on-line marketing. In most cases, scientists have focused on user-generated contents, which are freely accessible through different on-line services. Among such services, the incumbent one is indeed YouTube. This online platform was launched in 2005 and it currently features more than 6 billions hours of video watched every month (almost one hour per person on Earth), with more than 100 hours of videos uploaded every minute and 1 billion unique users per month. In order to analyze or predict content popularity, statistics about viewers, watch time and shares must be retrieved. The YouTube APIs, however, do not allow third parties to retrieve such an information in an open and accessible way. In order to overcome this problem, we have developed a framework, based on Web scraping techniques and big data tools, for the collection and analysis of YouTube video content popularity at scale. Our framework, called YOUStatAnalyzer, enables researchers to create their own dataset, according to a number of different search criteria and analyse them to extract relevant features and significant statistics.

YOUStatAnalyzer: a Tool for Analysing the Dynamics of YouTube Content Popularity

D. Miorandi;Francesco De Pellegrini
2013

Abstract

Understanding the dynamics of on-line content popularity is an active research field with application in sectors as diverse as media advertising, content replication and caching and on-line marketing. In most cases, scientists have focused on user-generated contents, which are freely accessible through different on-line services. Among such services, the incumbent one is indeed YouTube. This online platform was launched in 2005 and it currently features more than 6 billions hours of video watched every month (almost one hour per person on Earth), with more than 100 hours of videos uploaded every minute and 1 billion unique users per month. In order to analyze or predict content popularity, statistics about viewers, watch time and shares must be retrieved. The YouTube APIs, however, do not allow third parties to retrieve such an information in an open and accessible way. In order to overcome this problem, we have developed a framework, based on Web scraping techniques and big data tools, for the collection and analysis of YouTube video content popularity at scale. Our framework, called YOUStatAnalyzer, enables researchers to create their own dataset, according to a number of different search criteria and analyse them to extract relevant features and significant statistics.
978-1-936968-48-0
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11582/315127
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
social impact