A Data Exploratory Methodology To Understand The Users' Interactions In Nonlinear Web Multimedia Services

Nowadays, streaming companies, such as YouTube and Netflix, have prospered in Web multimedia market. In addition to presenting creative and engaging content, these companies have attracted users with a new standard to watch videos. Contrary to the traditional model, these services have used a nonlinear model, which users decide what, when and where to watch contents. Unconsciously, this user autonomy has required more quality of service for Content-Producers (CPs) and Management-Content Providers (MCPs) companies. While CPs have to produce more relevant contents and match them with users' tastes, the MCPs aim to ensure these contents reach the users with the highest quality. However, the dynamic behavior of end-users makes techniques inefficient over time. Thus, we present a data exploratory methodology in order to improve their services with the information extracted. It is organized into User Analysis, Content Analysis, and User-Content Interaction Analysis. We apply our methodology in real data from Samba Tech, the leader MCP in Latin America, and observe several possibilities to raise CPs' revenue by acquiring new customers and also to MCPs to improve mechanisms to reduce their costs.