Feb 22, 2016
Predicting YouTube Content Popularity via Facebook Data: A Network Spread Model for Optimizing Multimedia Delivery
Speaker : Yu-Hui ChenAuthors : Dinuka A. Soysa, Denis Guangyin Chen, Oscar C. Au, and Amine BermakFrom : 2013 IEEE Symposium on Computational Intelligence and Data Mining (CIDM)
Predicting YouTube Content Popularity viaFacebook Data: A Network Spread Model forOptimizing Multimedia DeliveryoutlineIntroductionMethodologySimulation resultsFuture workConclusion1.IntroductionThrough websites such as Facebook and YouTube to share multimedia content, the limited network resources, access to large amounts of multimedia data is a major challenge.
This paper proposes a Fast Threshold Spread Model (FTSM) to predict the future access pattern of multi-media content based on the social information of its past viewers.
FacebookYouTube,FTSM
32.MethodologyAn example infection process of Independent Cascade Model
Original infected nodes: Each infected node has a single chance of infecting his uninfected neighbors with a certain probabilityNewly infected nodes pass on the infection until no new nodes are infected4A) Facebook Data MiningExperimental setup: Requesting, downloading and analyzing JSON objects from Facebook
FacebookJSONJSON: Repeat for All pages of Each Friend: Parsers:Extracts specific data
5B) YouTube Video Statistics MiningThe YouTube statistics provided by YouTube API
YouTube APIYouTube6C) Fast Threshold Spread ModelG=(V,E)W(m)=0.5A1(m)+0.5A2(m)
D) Complexity Analysis on a Small Network vs a Large Network
3.Simulation resultsA) Determining Global ThresholdEffect on NumActiveNodes by changing the Threshold
B) Power Law behavior of the Facebook DatasetPlot of Node Degree vs Number of Nodes in linear scale
Node Degree :Number of Nodes:13B) Power Law behavior of the Facebook DatasetPlot of Node Degree vs Number of Nodes in log scale
14C) Correlation between Facebook social sharing and YouTube Global hit-countScatter plot of top 10 viral videos Global YouTube hit count vs FTSM predictors spread count
YouTube global hit count:YouTubeFTSM final spread value:FTSM15D) Transient spread simulation compared with YouTube dataNormalized view count for FTSM simulation (in red) and YouTube data (in blue) for top 9 viral videos in the Facebook Dataset
164.Future workFTSM for a large network of a few million nodes results in very long execution time.This paper is able to show that a small networks. A large network can be partitioned into multiple small networks .(ex. Hong Kong)FTSM
175.ConclusionThe Fast Threshold Spread Model (FTSM) was used to perform fast prediction of multi-media content propagation based on the social information of its past viewers.
This can be a solution to the cache management challenges when prioritizing.FTSM
18