Joint Session between Social Web Communities

1
Joint Session between Social Web Communities
Virtual games, interactive hosted services and
user-generated content in Web 2.0
2
Social Web Communities

• Core Question (compare Web page)
• How do resource sharing systems support knowledge
  sharing?
• New modes for knowledge sharing can be
  established on the Web?
• WG Mixed physical and Web communities,
  interlinking
• Which other types of information (beside just
  tags) would increase the usefulness of knowledge
  sharing on the Web?
• WG Community Dynamics I Hypothesis of efficient
  participation
• WG Community Dynamics II What do community
  structures evolve and contribute?
• How can methods and techniques from knowledge
  engineering allow a minimal-invasive use?
• WG Tag semantics

RealWorld
Participation Activity
Community Evolution
Tag Semantics
3
RealWorld
4
Conditions for the Real World Communities
RealWorld

• Linking data
• Interlinking data
• Identity
• Filtering by time, space, trust

5
Issues for Discussion
RealWorld

• Privacy
• Privacy protection and private sphere (right to
  be left alone)
• Put your email out, but do not be approached by
  unintended audience
• Fuzzy borderline between private and public data
  (intended use)
• Linked Data
• Wikipedia -gt DBPedia
• Licensing issues

6
Dunbar number
Participation Activity
Hypothesized cognitive upper limit to the number
of individuals one can form a social relationship
with at a given time. One of a series of
circles of intimacy' as hypothesized by Dunbar
to apply to human social relationships, where he
hypothesized that the number and quality of
relationships people have follow a vaguely
exponential curve, where one in general has 5
intimate friends, followed by 12-15 members in a
sympathy group, followed by 150 friends one can
maintain, followed by 1500 acquaintances.Origina
lly extrapolated by Dunbar from the size of the
human neo-cortexA study of Christmas Cards
distribution found a mean network size of 153.5
with the large deviation of 84.5 - (Hill and
Dunbar 2003)
7
Preliminary results 2
Participation Activity

8
Issues for joint discussion
Participation Activity

• ??

9
Communities Static Picture
Community Evolution
analyse or recommend
Research Community explicit (mailing list) /
implicit
?
Groups
Data mining
Social Network Analysis
Network based on similarity based on tags or
words (in papers)
Communication networks (email, )
tags, words
Folksonomy
Researchers
10
Evolution of Groups
Community Evolution
Communities Dynamic Picture
T1
T2
T3
Social network A
11
Possible Issues for Joint Discussion
Community Evolution

• Social implications of analysis outcomesyou
  are not important
• Which Web2.0 support would the user like to have
  or not to have
• Researcher survey quality indicators
  recommendations
• Tourist recommendations,
• Social risk management

12
Tag Semantics
Tag Semantics
Top 1000 tags from delicious 2006 crawl
13
Tag Semantics
14
Integrating Folksonomy Resources
Tag Semantics
15
Tag Semantics
Del.icio.us resources
http//theappleblog.com/2008/06/22
http//secrets.blacktree.com/
http//www.apple.com/macosx/
leopard
?
leopard
Flickr Photos
16
Issues for Joint Discussion
Tag Semantics

• Cross-linking between Platforms
• Cross-platform profiles (MyTag Personomies)
• Cross-platform content generation and usage
• Who is allowed to integrate?
• MyTag is a server
• Client-based infrastructure possible, too!?
• Colleague (Speck) from Kaiserslautern received
  warning letter from StudiVZ for doing analysis!

17
Panel Social Web Communities 5 Years into the
Future

• Bettina Hoser, Karlsruhe Fewer social
  networking platforms, but used daily
• Marko Grobelnik, JSI Hosting and analysis
  becomes easy
• Peter Mika, Yahoo Privacy is gone
• Alex Löser, TU Berlin Money, money, money -
  Efficiency, security, fun
• Ian Mulvany, Nature Publ.Things will not change
  that much that fast

18
Further Issues for Joint Discussion

• Further dimensions
• Incentives
• economics

19
Summary of Issues

• Cross-linking between Platforms
• Colleague (Speck) from Kaiserslautern received
  warning letter from StudiVZ for doing analysis!
• Social implications of analysis outcomesyou
  are not important
• Social risk management

• Privacy
• Privacy protection and private sphere (right to
  be left alone)
• Put your email out, but do not be approached by
  unintended audience
• Fuzzy borderline between private and public data
  (intended use)
• Linked Data
• Wikipedia -gt DBPedia
• Licensing issues