Visualizing Social Bookmarks

1
Visualizing Social Bookmarks
SIRTEL 2007

• Joris Klerkx, Erik Duval

http//www.cs.kuleuven.ac.be/hmdb/infovis/delicio
us/ www.cs.kuleuven.be/jorisk
2
Overview

• Situation
• Cluster Map Visualization in detail
• Demo
• Underlying Technology
• Challenges, Enablers

3
Social Bookmarks

• From bookmarks in the browser client
• to social bookmarks online on the Web
• Users classify content on the web by tagging
• These tags and bookmarks are public

4
Tagging Classifying Content

• Bottom-up unstructured approach
• folksonomies
• Top-down structured approach
• taxonomies, thesauri, ontologies

5
Tagging Problem

• They produce a flat namespace, rather than
  hierarchical structure of formal classification
  systems...
• But what about rich implicit structures between,
  tags bookmarks and users?
• Try to visualize these structures...

Hammond, T., Hannay, T., Lund, B, Scott, J.
Social Bookmarking Tools (I) A General Review.
D-Lib Magazine, 11(4), April 2005,
http//dx.doi.org/10.1045/april2005-hammond
6
Social bookmark tools

• Blue Dot
• del.icio.us
• Digg
• Diigo
• Furl
• NetVouz

• GiveALink.org
• My Web
• Reddit
• Simpy
• StumbleUpon
• etc.

• Many Users
• Lots of data
• Very Easy API to access data

7
Design Requirements (1/2)

• System should identify and visualize the tag and
  community structures
• gt 2 tags/users describe/share common bookmarks
• Users can become aware of these structures
• end users discover new bookmarks that could be
  of interest to them
• researchers experts study social bookmark data

8
Design Requirements (2/2)

• Use a novel paradigm instead of hypertext
  paradigm on del.icio.us website
• Enable users to explore social bookmarks in a fun
  and engaging space.

9
del.icio.us data

• gt 0 user del.icio.us usernames,
• load the tags and bookmarks of
• these user(s)
• the users in the network of these user(s)
• of the users that are a fan of these user(s)

is fan of
10
Cluster Map Visualization

• Consists out of 3 parts
• Tree Structure
• Cluster Map Visualization
• Result list

1
2
3
11
Tree Structure

• Overview of ALL tags and users
• Currently loaded into the system
• Number of matching bookmarks
• User Interaction
• Checkbox selection
• Sort
• Filtering box

12
Cluster Map (1/3)

• Bookmark(s)
• small circles
• belongs to gt 0 user(s)
• User(s),
• big circles with
• name icon
• number of bookmarks
• Clusters of bookmarks
• circles between user(s)
• if selected, adapt numbersof user(s)

13
Cluster Map (2/3)

• Tags, e.g. web
• big circles with
• name
• number of bookmarks
• Clusters of bookmarks
• circles between users tags
• if selected, adapt numbersof users and tags

14
Cluster Map (3/3)

• User Interaction
• context menu(s)
• Users
• Add all data
• Add to own del.icio.us network
• Posts
• Post to del.icio.us

15
Result list

• Search
• within loaded data
• Detailed result metadata
• title, tags, users, location
• User Interaction
• add tags, users to visualization
• open in browser
• post to del.icio.us

16
Typical Use

• Start from egocentric point-of-view
• Start with you know, then grow
• Less perceptual and computational burden
• so..
• load all data from a chosen username
• tags, bookmarks, network, etc.
• perform query, select tags or users in tree view
  and add items step-by-step to visualization
• Explore visualization and thus del.icio.us data

17
Demo
Try-it yourself!
http//www.cs.kuleuven.ac.be/hmdb/infovis/delicio
us/
18
Information Visualization Framework

• Open Extensible
• Plugin different visualization techniques
• treemaps, hyperbolic trees, node-graphs, fisheye
  views, clustermaps, etc.
• Create different case studies
• Del.icio.us, Ariadne KPS, Alocom repository,
  Calibrate Project

19

• del.icio.us social bookmark data as a treemap

20

• Social bookmark data of the Calibrate Project
• Question are there common bookmarks between
  users? yes..

21
Future Work

• del.icio.us Tool
• Evaluation gather user data, usability tests,
  etc.
• Add functionality
• Framework
• Other datasets
• Other visualizations

22
Challenges, Enablers

• Challenge If these kind of visualization tools
  will be used, depends on
• the kind of user, eg. experts versus normal user
• the benefit of these kind of tools
• Enablers
• Make visualization tools fun, playful and
  engaging, fitting in the workflow for end users
• keep track of what users actually do and adapt to
  match the needs

23
Thanks for your attention!
SIRTEL 2007
Questions?
Try-it yourself!
http//www.cs.kuleuven.ac.be/hmdb/infovis/delicio
us/
24
Have a del.icio.us lunch!
SIRTEL 2007