270 likes | 295 Views
This survey explores the diverse user behavior in online social networks, covering activities like friendship creation, content publishing, messaging, and more. It delves into connectivity, interaction, traffic activity, and mobile social behavior to enhance user experience and address security concerns.
E N D
Understanding User Behavior inOnline Social Networks: A Survey Long Jin, University of California, San Diego Yang Chen, Duke University Tianyi Wang, Tsinghua University Communications Magazine, IEEE, 2013, 51(9) PresentedByTongShensi 2015.11.12
Introduction • Connectivity and Interaction • Traffic Activity • Mobile Social Behavior • Conclusions
Introduction • Motivation • OnlineSocialNetworks(OSN)havedramaticallyexpanded • OSNuserbehaviorcoversvarioussocialactivitiesthatuserscandoonline • Friendshipcreation • Contentpublishing • Profilebrowsing • Messaging • Commenting
Introduction • Motivation(cont.) • OSNuserbehaviorisimportant • Internetserviceproviders • OSNTrafficisgrowingquicklyandbecomingsignificant • Guidethemtodosomeinfrastructuralaction • OSNserviceproviders • Understandingcustomers’attitudetowarddifferentfunctions • Understandingusers’geographicdistributionandtrafficacitivity • OSNusers • Enhanceuserexperience • Blockingmalicioususers
Introduction • Organization • Connectivityandinteraction • SocialgraphcanrepresentrelationshipbetweenusersinOSNs • HasbeenwidelyusedinOSNresearch • Trafficactivity • UnderstandthenetworkusageofOSNs • Mobilesocialbehavior • Enhancetheperformanceofmobilesocialapplicationsandsystems • Maliciousbehavior • Security&Privacy
Introduction • Connectivity and Interaction • Traffic Activity • Mobile Social Behavior • Conclusions
ConnectivityandInteraction • Motivation&Challenges • SocialgraphcanrepresentrelationshipsamongusersinOSNs • Types • Undirectedgraphs • Directedgraphs • Thehugesizeofsocialgraph • Samplingandcrawlingtechiques
ConnectivityandInteraction • Solution&Discussion • UndirectedGraphModel • Everyuserisdenotedasanode • Friendshipbetweenanyuserpairisrepresentedbyanedge • Wilsonet al. foundthatuserstendtointeractwithonlyasmallsetoffriends • Onlyvisibleinteractionbetweentwousercreateanedge • Laterperformsbetter
ConnectivityandInteraction • Solution&Discussion(cont.) • DirectedGraphModel • Jiang et al. studied latent graph • Renren tracks the most recent nine visitors to every users’ profile • A directed edge from A to B indicates A has visited B’s profile • Prevalent and frequent than visible intersection • Uncorrelated with the frequency of content updates or number of friends
ConnectivityandInteraction • Solution&Discussion(cont.) • DirectedGraphModel(cont.) • Hwak et al. studied Twitter’s graph • A directed edge from A to B indicates A has subscribed to receive B’s latest news • Basic information overview of Twitter • Distribution of followers/followees • analyzes how the number of followers or followees affects the number of tweets
ConnectivityandInteraction • Solution&Discussion(cont.) • Graph Sampling • A fast increase in the number of users • Make the size of social graphs larger and larger • Challenge performing any analysis with limited computation and storage capability • Graph sampling • Preserve the origin graph’s property • Breadth-First Sampling(BFS) • Random Walk(RW)
ConnectivityandInteraction • Future Work • Dynamic feature • Much of existing work study in a relatively static way • Dynamic feature could deeply understand OSN’s user behavior • Like new users join OSNs, make new friends…
Introduction • Connectivity and Interaction • Traffic Activity • Mobile Social Behavior • Conclusions
Traffic Activity • Motivation & Challenge • Graph contains limited information • Can interpret how users use OSNs better • For ISP, they have strong incentive to get better understanding of how the traffic pattern between end users and OSN sites will evolve
Traffic Activity • Solution & Discussion • Traffic Monitoring • Benevenuto et al. analysis user behavior based on detailed clickstream data • The frequency of accesing OSNs • Total time spent on OSNs • Session duration of OSNs • Silent or latent Interactions such as browsing account for more than 90 percent of user activity
Traffic Activity • Solution & Discussion • Traffic Monitoring(cont.) • Schneider et al. also study clickstream data • But focus on ISPs aspect • Like which features account for most traffic bytes
Traffic Activity • Solution & Discussion • Locality of Interest • Facebook is heavily dependent on centralized US data center • Slow response time & unnecessary traffic • Wittieet al. analysis these two problems • Partitioning & distribution • 79 percent faster and 91 percent less bandwidth
Traffic Activity • Solution & Discussion • Navigation Characteristic • Dunn et al. try to understand the similarities and differences in the web sites users visit through OSNs vs. through search engines. • OSN visitors are less likely to navigate to external web sites • OSNs direct visitors to a narrower subset of the web than search engines
Traffic Activity • Future Work • Most existing analysis are led by either academic groups or ISPs, without OSN service provider • Academic groups use extensive crawling to obtain data, which encounter many restrictions • ISPs can only get a partial view of the whole site • Envision that OSN providers can collaborate with researchers in order to understand user behavior in an insightful way
Introduction • Connectivity and Interaction • Traffic Activity • Mobile Social Behavior • Conclusions
Mobile Social Behavior • Motivation & Challenge • More and more OSN services have been expanded to mobile platforms • More mobile-centric functions have been integrated into OSNs • Understanding mobile social networks(MSN) user behavior is very helpful for the design and implementation of MSN systems
Mobile Social Behavior • Solution & Discussion • Mobile Social Application • Calculates similarity score to recommend nearby friends • Geographical Prediction in OSN • Predict a users location according to his/her friends’ location • Friendship and Mobility in LBSN • Analysis the relationship between friendship and human movement
Mobile Social Behavior • Future Work • There are several fundamental issues that require continuous exploration in the research related to user behavior in MSNs • Social data delivery and social applications in mobile environments rouse challenges in several layers of the Internet protocol stack
Introduction • Connectivity and Interaction • Traffic Activity • Mobile Social Behavior • Conclusions
Conclusion • Study user behavior in OSNs from four different perspectives • Connection and interaction • Traffic activity • Mobile social behavior • Malicious behavior • Will enhance the user experience from various aspects • We believe future research will generate more interesting research problem