Improving Content Delivery on User Behavior Using Data Analytics

Abstract

Internet traffic has increased, mainly as video viewers have increased. Business estimates indicate that video accounts for 90% of Internet traffic by 2020. Due to the rising traffic, content distribution is underlined by the growing volumes of video traffic. The content delivery model must be planned, delivered and managed adequately to meet these rising requirements. We identified many user access trends with large-scale analysis of user behavior data which have significant consequences for the design in our unique dataset, including 100 video selections from the two largest Internet video providers spanning about two months. These involve a partial emphasis on material, regional interests, transitional times and trends. By conducting a comprehensive measurement analysis, we examined the effect of our results on the designs. There is considerable synchronous viewing behavior for video content, which may increase video federation availability considerably by as much as 95%.