An Engagement Model Based on User Interest and QoS in Video Streaming Systems

Abstract

With the surging demand on high-quality mobile video services and the unabated development of new network technology, including fog computing, there is a need for a generalized quality of user experience (QoE) model that could provide insight for various network optimization designs. A good QoE, especially when measured as engagement, is an important optimization goal for investors and advertisers. Therefore, many works have focused on understanding how the factors, especially quality of service (QoS) factors, impact user engagement. However, the divergence of user interest is usually ignored or deliberatively decoupled from QoS and/or other objective factors. With an increasing trend towards personalization applications, it is necessary as well as feasible to consider user interest to satisfy aesthetic and personal needs of users when optimizing user engagement. We first propose an Extraction-Inference (E-I) algorithm to estimate the user interest from easily obtained user behaviors. Based on our empirical analysis on a large-scale dataset, we then build a QoS and user Interest based Engagement (QI-E) regression model. Through experiments on our dataset, we demonstrate that the proposed model reaches an improvement in accuracy by 9.99% over the baseline model which only considers QoS factors. The proposed model has potential for designing QoE-oriented scheduling strategies in various network scenarios, especially in the fog computing context.

Predicting User Quitting Ratio in Adaptive Bitrate Video Streaming

“…One novelty in this paper can be easily identified in the special case of initial buffering, which was tested in experiments D and E and is marked in Figure 5 by lines segment. It can be seen that user drop increases non-linearly with the initial buffering duration (as expected from past studies [26], [28]), but large differences between experiments D and E are found. Indeed, in experiment E, users reacted more sharply to initial buffering than in experiment D due to the differences in total stalling duration between experiments D and E (272 and 341 s, respectively).…”

“…Second, stalling midway through a video is discussed. In this case, abandonment is found to increase logarithmically with the stalling duration [26], [28]. Going from 12 to 24 s results in abandonment increasing by 30% [29].…”

“…For example, for the same total stalling duration, abandonment was found to increase by a factor of 3 if the number of stalling events goes from 2 to 3 [30]. Another example is rebuffering ratio (total stalling duration divided by playback duration) studies that showed that increasing the buffering ratio results in increasing abandonment [27], [28], [31]- [33] (a 1% increase in the rebuffering ratio reduced viewing time by 67% [27], a 0.13 increase in stalling per minute decreased viewing time by 20 min [32], and an 1% increase in the rebuffering ratio results in a 3.33% reduction in viewing time [34]).…”

Quitting Ratio-Based Bitrate Ladder Selection Mechanism for Adaptive Bitrate Video Streaming

“…Then, another significant body of work worth mentioning is the studies of the relationship between viewing time and quality impairments such as initial loading delay [40]- [42] and midway-though stalling [43], [44]. These studies showed that the odds of users quitting increase exponentially as they need to wait for the video to play.…”

When Should Recommenders Account for Low QoS?

“…In [41], [42] the authors considered user engagement as their target QoE metric. In particular, in [42] the authors accounted for users' interests and QoS factors to build an engagement/QoE predictive model. Our approach is aligned with this integrated perspective to predict QoE.…”