Developing a predictive model of quality of experience for internet video

Abstract

Improving users' quality of experience (QoE) is crucial for sustaining the advertisement and subscription based revenue models that enable the growth of Internet video. Despite the rich literature on video and QoE measurement, our understanding of Internet video QoE is limited because of the shift from traditional methods of measuring video quality (e.g., Peak Signal-to-Noise Ratio) and user experience (e.g., opinion scores). These have been replaced by new quality metrics (e.g., rate of buffering, bitrate) and new engagement-centric measures of user experience (e.g., viewing time and number of visits). The goal of this paper is to develop a predictive model of Internet video QoE. To this end, we identify two key requirements for the QoE model: (1) it has to be tied in to observable user engagement and (2) it should be actionable to guide practical system design decisions. Achieving this goal is challenging because the quality metrics are interdependent, they have complex and counter-intuitive relationships to engagement measures, and there are many external factors that confound the relationship between quality and engagement (e.g., type of video, user connectivity). To address these challenges, we present a data-driven approach to model the metric interdependencies and their complex relationships to engagement, and propose a systematic framework to identify and account for the confounding factors. We show that a delivery infrastructure that uses our proposed model to choose CDN and bitrates can achieve more than 20% improvement in overall user engagement compared to strawman approaches.

Predicting the Quality of Experience for Internet Video with Fuzzy Decision Tree

“…Any of these factors may lead to different user behavior patterns and further affect both user engagement and user QoE for Internet video. In other words, observations made in [5] may not be coincident with the environment of Internet video industry in China. Meanwhile, China owns a very large amount of Internet video users, as described earlier.…”

“…Over the past few years, there have been several studies on QoE for Internet video [1,3,4,5]. From the observations made by these researches, we can draw the conclusion that video quality influences user engagement.…”

“…Furthermore, confounding factors such as content type and region may affect both the quality metrics and user engagement. In [5], Balachandran A et al analyze the complex relationship between quality metrics and user engagement, the interdependence between each quality metric and the influence of confounding factors. To tackle all these challenges in developing a actionable and accurate predictive model for user QoE during the video watching, they run machine learning algorithms on a dataset that captures large-scale measurements of video quality and user engagement.…”

“…Our work in this paper is based upon theirs. However, there are two main differences: (1) Our data is gained from one of the most popular video service providers in China and the users of it are mainly Chinese, while the study in [5] is limited to viewers in America. There are a lot of distinct differences between China and other countries in the state of network development, the user group characteristics, the cultural background and many other aspects.…”

Improving Energy Efficiency for Mobile Media Cloud via Virtual Machine Consolidation

“…On the contrary, the transcoding service is a computation-intensive task, but it needs little bandwidth. Therefore, a natural question arises: How to achieve the optimal CPU and bandwidth utilization simultaneously with minimal energy cost is of critical importance [1]. Fortunately, modern virtualization technology makes it possible to save operational costs such as power by VM consolidation [6,9].…”