Exploring the Use of Tags for Georeplicated Content Placement

Delbruel, Stéphane; Frey, Davide; Taı̈ani, François

doi:10.1109/ic2e.2016.37

Cited by 1 publication

(3 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because the viewing patterns of many videos in UGC services present clear geographic trends [7], which are strongly correlated with a video's tags [11]. Table 1 shows for instance how the tags "bollywood " and "favela" follow clearly distinctive geographic distributions in a Youtube dataset analyzed in an earlier work [11]. Correctly predicting the geographic distribution of a video's views is particularly important in decentralized systems that often lack the caching infrastructure of large integrated services.…”

Section: Placing New Videos: the Prediction Problemmentioning

confidence: 99%

“…Here for instance, Alice has viewed a BBC video with the tag 'news' ( ), and a video on environmental protection with the tags 'news', and 'animals' ( ). We rely on a tag-based affinity function f that measures a user's affinity with new videos (5) [5,11]. The only assumptions we make about f is that its result is correlated with the probability that this user will watch the video (6).…”

Section: Problem Statement and Related Workmentioning

confidence: 99%

“…We have shown in a previous work that the tags attached to videos are a good predictor of a video's view distribution [11]. Unfortunately, individual peers do not by default have access to the past videos and tags consumed within individual countries, and this information can be costly to aggregate explicitly.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Mignon: A Fast Decentralized Content Consumption Estimation in Large-Scale Distributed Systems

Delbruel¹,

Frey²,

Taı̈ani³

2016

Distributed Applications and Interoperable Systems

Self Cite

View full text Add to dashboard Cite

International audienceAlthough many fully decentralized content distribution systems have been proposed, they often lack key capabilities that make them difficult to deploy and use in practice. In this paper, we look at the particular problem of content consumption prediction, a crucial mechanism in many such systems. We propose a novel, fully decentralized protocol that uses the tags attached by users to on-line content, and exploits the properties of self-organizing kNN overlays to rapidly estimate the potential of a particular content without explicit aggregation

show abstract

Section: Placing New Videos: the Prediction Problemmentioning

confidence: 99%

Section: Problem Statement and Related Workmentioning

confidence: 99%