Nowadaysonline networking sites provide user platforms to interact with each other. Communicating and sharing information in onlinesocial networks such as Facebook, twitter or flicker are an important part of peoplesever day life, which leads to huge amount of data each day.
The importantquestion is how to use this data in order to extract useful information. Flickeris an online photo and video sharing platform with social networking features,founded in 2004. In 2008, Flicker claims to have more than 3 billion images.Flicker allows users to create friendship and add their favorite photos totheir profiles.Oneof the amazing features of social networks is how they play a fundamental role inInformation propagation. Word of mouth is widely been used as a marketinginstrument to spread content and ideas about products through population widelyand quickly. In this paper authors try to answers questions about how widely andquickly information spread in social networks? Is this information spreading locallyor globally? For example if an image is popular in a certain region will it beadded as favorite in different parts of the network? How quickly informationpropagation happens? Do people discover their favorite photos through theirlinks (friends)? Inorder to answer these questions Cha et al.
collect the information of flickerfrom 2.5 million users and 33 million links between and they try to capture thedynamics of information propagation for 104 consecutive days. First theyrandomly select users and by following their friends links to get a “snowball”sample of the Flickr social network. The authors examine different network structuralproperties such as nodes in and out degrees, diameter, path length, ClusteringCoefficient, etc.
The flicker networks exhibits a small-world network structurewhich implicates that information spreading could happen in flicker over shortnetwork paths.Inthis paper authors focus on the fan popularity of pictures and try to investigatehow fast and widely users add photos as their favorites. First they compare themost popular local and global list of photos while they assuming that if photoswidely spread throughout the networks there will be a high similarity betweenthe two. They discover no overlap between the local and global hotlists in theone-hop neighborhood which shows strong locality across most of the popularitylevels. Later authors investigate how the distances from the uploaders affectthe distribution of fans of photos. By taking into account the different informationpropagation mechanisms through flicker (Search results, External links, Word-of-mouthetc.
), the authors find that Social cascade playsan important role in information spreading through Flicker. Cha et al do not considerFlicker as a general online social network and suggest apply the same research togeneral cases and find out if the results changes for much larger networks?