Dekscom EyeNet Video Big Data Analysis System (EVQBD) collects massive user viewing behavior and viewing quality data through tens of millions of QoS agents embedded in the IPTV/OTT/DVB video terminal, and then imports the data into Spark &Hadoop-based big data analysis system. EVQBD system performs real-time and offline statistical analysis of massive data using Spark+HIVE+HDFS-based big data engine, and provide data visualization through flexible custom reports and diversified charts.

System Key Functions and Applications

The main applications and values of the EVQBD system includes

•     Live channel rating and on-demand hot program statistics: based onmassive data collection and analysis froma large number of video users, the live channel rating and the on-demand hot program statistical analysis can be more accurately performed, which provides a good reference for the video advertisement delivery and operation.

•     Refined video content operation can be realized by analyzing the user’s viewing preferences, operating characteristics and performing user portraits, and the reference basis for the user’s personalized charging programs, on-demand programs and other application recommendations can be provided.

•     IP network and CDN performance optimization: By performing multi-dimensional KPI statistical analysis on the viewing quality data of video users, it can provide a powerful reference for IP network and CDN performance optimization, and improve the overall user experience quality.

•     Efficient video user management: By tracking the trend of new users and analyzing the development of loyal users, active users, inactive users, and lost users, efficient user management can be achieved, and operational support can be more effectively coordinated with marketing activities.

Live channels rating and statistics

On-demand hot programs statistics