This document discusses work package 4 (WP4) of a project which involves developing extensions to existing collaboration tools to support diversity. It outlines two tasks: T4.1 which involves developing extensions for tools like WordPress and MediaWiki over three years, and T4.2 which involves producing best practice documents. It then provides more details on the planned extensions for WordPress and MediaWiki, and brainstorms potential scenarios for applying the extensions, such as adding links to related but different opinions in blogs, and checking biases in sources for wiki articles.
2. Project overview
WPs 6: Dissemination, exploitation, community
WP5: Case studies
WP7: Project management
WP4: Diversity toolkit
Media Wiki Yammer WordPress Google Wave
WP3: Diversity representation &processing
building
Models and Search, selection Presentation
Summarization
metaphors and ranking and interfaces
WP2: Diversity mining
Opinion
Multilinguality Bias in media Fact coverage Story links
detection
WP1: Data collection and management Annotation,
integration,
Blogs News Twitter Wikipedia LOD linking
3. WP Overview
o 2 Tasks
o T4.1: Extensions to Web 2.0 based collaboration platforms
• One deliverable at end of each project year
o T4.2: Best practices (M18-M36)
• One deliverable at end of second and third year each
4. T4.1: Extensions to existing tools
o First set of extensions, including WordPress and MediaWiki
o Evaluation of candidates at M12 for further tools
o Candidates: Google Wave and Microblogging
o Based on expected impact and case study needs
(especially Telefónica)
• Which ones would that be?
5. WordPress
o Most popular CMS in use today
o Used by 12% of the top 100k Websites
o 25 Mio+ installations
o Grew from a blogging systems
o Full-fledged CMS by now
o Huge amounts of extensions
o PHP
o GPL
6. MediaWiki
o Most popular wiki on the Web today
o Developed for Wikipedia, but was generalized
o Still geared towards high performance
o More than 1000 extensions, like Semantic MediaWiki
o PHP
o GPL
7. Scenario brainstorm I
Target: Bloggers as early adopters
o Add links to related, but different opinions
Requirements:
• Topic discovery on the blog item
• Opinion / Bias detection on the item
• Finding items on the same topic with different opinions
8. Scenario brainstorm II
Target: Wiki-editing
o Check referenced sources
• For every source, check how it is biased
• Find how many sources are biased in order to predict the
bias of the overall article
• Evaluate if the source bias translates to content bias
o Discover editor bias
• To what group of bias does an editor belong to
• Does past behavior predict future behavior?
9. Scenario brainstorm III
Target: Semantic MediaWiki and diversity
o Shortipedia (demo)
o Anyone can claim anything about anything, but source
o Mark up source with the exact place in the text for the claim
• Train learning
• Discovering user bias and bias history of an article
• evaluating if a claim still is sourced
o Representations for sourced claims
• Reification, name graph?
10. Scenario brainstorm IV
Target: Accessing Forum information
o Access for Telefónica
o Discover Hotspots in Forum discussions
o Discover interestingness of singular postings in order to
guide the user for reading