Measurement-driven Modeling and Design of Internet

Download Report

Transcript Measurement-driven Modeling and Design of Internet

Enabling the Social Web
Krishna P. Gummadi
Networked Systems Group
Max Planck Institute for Software Systems
My research
• Understand and build complex networked systems
• Examples:
– social web systems: e.g., Facebook, Twitter, YouTube
– Internet access networks: e.g., cable, DSL broadband
– peer-to-peer systems: e.g., BitTorrent, Skype
• Aspects of their complexity
– massive scale
– tremendous heterogeneity
– decentralized control
My methodology
• First understand and then build
– observe deployed systems
– extract feedback
– test new designs and architectural principles
• Why understand?
– Can’t predict overall system behavior from first principles alone
• much like social, economic, or political systems
The big picture
Three fundamental trends & challenges in social Web
1. User-generated content sharing
–
can we protect privacy of users sharing personal data?
2. Word-of-mouth based content exchange
–
can we understand & leverage word-of-mouth better?
3. Crowd-sourcing content rating and ranking
–
can we find trustworthy & relevant content sources?
Challenge:
Privacy concerns with personal data sharing
• The traditional web (1993 – )
– Publishers: Companies, Universities, & Governments
– Content: Public information
– Openness & universal access were key goals
• The social web (2005 – )
– Publishers: Individuals
– Content: Private photos & videos
– Privacy and access control are key challenges
Research problems
• Data uploaded to social networking sites
– Can we use home network infrastructures to share
data?[NOSSDAV ’11]
• Data given to 3rd party social networking apps
– Can we use trusted cloud infrastructures to host apps?
[WOSN ’12]
• Data shared with other users in the network [IMC ’11]
– Can we design better access control mechanisms?
– Can we design abstractions to control data exposure?
• Data implicitly leaked by friends [WSDM ’10]
Challenge:
Understanding dynamics of word-of-mouth
• Discovering information on the Web
– Old method: Browsing from authoritative sources
– New method: Word-of-mouth from friends
• Lots of theories & beliefs about viral propagation
– But few are empirically derived or validated at scale!
• Large-scale empirical studies only possible recently
– Measurements of social network graphs, their evolution, & user
activity [IMC ‘07, UbiComp ’07, WOSN ’08, WOSN ’09]
Research problems
• Understand dynamics of propagation
– Temporal and spatial patterns of propagation
– Role of social network, social systems, and user influence
• For different types of information and innovations
– News, web URLs, conventions, and technology services
• With the ultimate goal of enabling better viral campaigns
– Consumers: Help them get content they would not otherwise receive
– Publishers: Help them spread their content more effectively
Studies of information diffusion
• How photos spread in Flickr [WOSN ’08, WWW ’09]
• How web URLs & news spread in Twitter [IMC ’11,
ICWSM ‘11]
• The role of influencers and offline geography in
information dissemination in Twitter [ICWSM ’10, ’12]
• Understanding and predicting the spread of social
conventions in Twitter [ICWSM ’12, CIKM ’12]
Challenge:
Finding relevant & trustworthy content
• Traditional web search leverages content-content links
• But, social web content is often multimedia & real-time
– No
links to other content
• Instead, the content is ranked collaboratively by users
• Concerns with relevancy and trustworthiness of users
– How to identify users with similar interests
– How to separate authoritative sources from spammers
Research problems
• Observation: Content is inter-linked
via social network
• Leverage user network & activities
Web
– To find people with similar interests?
– To isolate malicious users and
spammers?
• Concerns: anonymity & privacy
Social networks
Finding trustworthy & relevant users
• Built systems that leverage social network to
–
–
–
–
limit the amount of spam in communications [NSDI ’08]
limit large-scale data aggregation in social sites [CoNext ’12]
improve web search results [HotNets ’06]
detect or tolerate Sybil identities [SIGCOMM ’10, EuroSys ‘12]
• Finding authoritative & trustworthy users in Twitter
– understanding & combating link farming [WWW ’12]
– identifying & ranking topical experts [WOSN ’12, SIGIR ’12]