social networking for scientists (research communities) using tagging and shared bookmarks: a web...
TRANSCRIPT
Social Networking for Scientists (Research
Communities) Using Tagging and Shared Bookmarks: a
Web 2.0 ApplicationMarlon Pierce, Geoffrey Fox, Joshua Rosen, Siddharth
Maini, and Jong Youl Choi
Community Grids Lab, Indiana University
Collaborative Technologies and Systems, 2008
AbstractReport their investigation and implementationo Creating communities of researchers
Motivation is to provide outreach tools that broaden the participation of these groups in funded research activities
IntroductionEnabling researchers to find both useful online
resources and also potential collaborators on future research projects
MSI-CIEC is an NSF funded project to engage researchers at minority-serving institutions in modern cyberinfrastructureo “Minority Serving Institution-Cyberinfrastructure
Empowerment Consortium”
The MSI-CIEC social networking Web portal combines social bookmarking and tagging with online curricula vitae profiles
Enable MSI researchers to find others with similar research interests
Tag Clouds
Click-Taggable Grants.gov RSS Feed
Click a tag and see all associated links.
CapabilitiesThe portal is designed to support academic user
communitieso online user profileso shared online bookmarkso tags
CapabilitiesPublic user profiles with the user’s tag cloud
RSS feeds
Click tagging
Bookmark any URL and users describe bookmarks with one or more keyword tags
Users can search award funding and project data
User’s tag cloud
Social networking information
RSS feeds
Tagging
Intended to foster research collaborations
Interesting or uninteresting
NSF Tag Cloud
• The Tag Cloud displays all meta tags which were created automatically from the NSF Awards
• You can browse all the projects related to these meta tags
• Eg. Clicking “2007” would display all the people who have the tag “2007” meaning that they worked in projects awarded in the year 2007.
• Meta Tags:• Year name “2008”• Size of the Project “Small” “Medium” “Large”• NSf Directorate “GEO” “BIO” etc.
USER WHO HAVE TAG “2007”USER WHO HAVE TAG “2007”
Researcher’s tag cloud and list of funded projects
Tagging and Folksonomies
Exploring communities in Collaborative Tagging SystemsA user may want to see other people who
have tagged on the same object
Find a group of people who might have the same interest and look at their bookmarks or resources
Models of Collaborative Tagging SystemMain elements of collaborative tagging systems
consist of tags, recourses, and userso Tag can be keywords, terms, or neologismso Graphical connections
Models of Collaborative Tagging SystemGeneral purpose of tagging systemso Find specific resources tagged collaboratively by
multiple userso Retrieve information about resources or users
Models of Collaborative Tagging SystemTwo different models:o Vector space model
Frequencies of tag occurrences for searching Information retrieval
o Graph model Graphical characteristics• Path and degree of connectivity between nodes
Social network analysis
Models of Collaborative Tagging SystemVector space modelo A resource (or a user) is represented as a vector
of tagso For example:
A resource tagged by 2 occurrences of tag1, 1 occurrence of tag2• <2, 1>
Models of Collaborative Tagging SystemGraph modelo Searching is task to find specific properties in
the graph
Discovering CommunitiesMost common examples of social activities in a
network is expressing one’s interests
Finding a group of people who are working on the same topics or interests which we call “discovering a community”
Frequency Analysis and ClusteringFind information:o More frequently used tagso More referenced resourceso More actively involved users
Structural AnalysisThe structural analysis considers the tagging
activities as a grapho More intuitive and human-understandable
Help users to find other informationo Connectivity, connection distances between
userso Size of communitieso Degree of strength of a connection
Summary and future workThis paper describes the design and
implementation of the MSI-CIEC Networking Portal
This work is motivated by the need to support social networks of researchers
Using the portal as a laboratory for core computer science work on social network analysis
The key problem with most social network applications is the lack of interoperabiity