Analysis of user behavior in a large-scale internet video-on-demand(VoD) system
Yaohui Yuan, Xingjun Wang, Guangxiang Bin · 2020
With the development and popularity of the Internet, Internet traffic has increased dramatically. Numerous studies have shown that video accounts for a large percentage of Internet traffic, and this percentage is still rising in the future. A good understanding of user behavior in online video systems can help us design, configure and manage video content distribution to alleviate network stress. In this paper, we did a detailed analysis of user behavior data for Internet video. Our research shows that the user's daily access and online pattern of users have a fixed pattern, and the user's access behavior conforms to Zipf's law. Besides, we optimized the fit of Zipf-like distribution of video's popularity. Finally, we built a reliable simulation system that simulates user behavior data. Overall, we believe that the results presented in this paper are very important and valuable to the whole network.