Real-Time Video Content Popularity Detection Based on Mean Change Point Analysis

Abstract

Video content is responsible for more than 70% of the global IP traffic. Consequently, it is important for content delivery infrastructures to rapidly detect and respond to changes in content popularity dynamics. In this paper, we propose the employment of on-line change point (CP) analysis to implement real-time, autonomous and low-complexity video content popularity detection. Our proposal, denoted as realtime change point detector (RCPD), estimates the existence, the number and the direction of changes on the average number of video visits by combining: (i) off-line and on-line CP detection algorithms; (ii) an improved time-series segmentation heuristic for the reliable detection of multiple CPs; and (iii) two algorithms for the identification of the direction of changes. The proposed detector is validated against synthetic data, as well as a large database of real YouTube video visits. It is demonstrated that the RCPD can accurately identify changes in the average content popularity and the direction of change. In particular, the success rate of the RCPD over synthetic data is shown to exceed 94% for medium and large changes in content popularity. Additionally, the dynamic time warping distance, between the actual and the estimated changes, has been found to range between 20 samples on average, over synthetic data, to 52 samples, in real data. The rapid responsiveness of the RCPD is instrumental in the deployment of real-time, lightweight load balancing solutions, as shown in a real example.Index Terms-Video content popularity detection, change point analysis, on-line change point detection, binary segmentation algorithm, load balancing.

Real-Time Algorithms for the Detection of Changes in the Variance of Video Content Popularity

“…On the other hand, the problem of detecting (i.e., estimating), non-parametrically and in real time, CPs on content popularity sequences, has not been adequately investigated yet. Among of only a handful of related studies, in our previous works [11], [12], [44] we proposed and implemented a real-time, non-parametric and low-complexity video content popularity CP detector (as opposed to predictor) for changes in the mean value of video content popularity. In the present contribution, in contrast to [11], [12], we introduce an innovative online algorithm for the detection of CPs in the second order statistics of content popularity data.…”

“…We note that standard off-line CP schemes can only detect a single CP. To address the issue of detection of multiple CPs, we modify the basic scheme with a novel time series segmentation heuristic, that belongs to the family of binary segmentation algorithms, similarly to [11], [12].…”

“…Importantly, instead of attempting to predict the evolution of content popularity, in this work we rather focus on detecting changes in its underlying statistics, and doing so in real-time. To this end, we propose the use of on-line change point (CP) analysis; to complement our work [11], [12] that focused on the identification of changes in the mean of a time series, here, we alternatively investigate the performance of corresponding on-line algorithms to identify changes in the variance of a time series using CP analysis.…”

“…In this work, we propose an efficient combination of an off-line and various on-line procedures for the detection of changes in the second order statistics of video content popularity, as soon as they occur (real-time). The proposed detector is built upon our earlier proposal for a real-time CP detector of mean changes in data series, that we applied to monitor the average number of video content [11], [12]. Albeit, the monitoring of changes in the variance of a time series is much more challenging.…”

“…To identify changes in the variance, a more elaborate test statistic is employed in the present study. With respect to [11], [12], we further introduce novel on-line tracking mechanisms based on autoregressive moving average (ARMA) and generalized autoregressive conditional heteroskedasticity (GARCH) models. The most important novel aspects of this paper are listed below:…”