Optimization of file allocation for video sharing servers
NTMS'09 Proceedings of the 3rd international conference on New technologies, mobility and security
Cooperative caching versus proactive replication for location dependent request patterns
Journal of Network and Computer Applications
| Hi-index | 0.00 |
Online social media applications have exploded in popularity in the Web. In most of these applications, users interact with other users and create content that becomes available on the Web, such as textual information, photos, and videos. YouTube is the largest online social video sharing service, that generates a huge amount of Web traffic. This paper presents a geographical characterization of YouTube usage. It analyzes video and user characteristics for different geographical regions, concentrating mainly on Latin America. We develop efficient crawlers for collecting data about videos and users. Because the number of videos and users is very large, we sample from the object spaces, sampling over 2 million videos and 5 million users. Based on the collected data, we show that there exists relationships between geography and the social network features available in YouTube. We present evidences that indicate that geography creates a locality space in YouTube, which could be used to explore infrastructure improvements, such as caching, content distribution networks and broadband pricing mechanisms.