Proposal:Peer voting-based bias indicator for controversial topics
Every proposal should be tied to one of the strategic priorities below.
Edit this page to help identify the priorities related to this proposal!
- Achieve continued growth in readership
- Focus on quality content
- Increase Participation
- Stabilize and improve the infrastructure
- Encourage Innovation
|
Share this:
|
Summary
This approach offers a method for determining potential bias in a Wikipedia article through the use of controlled public voting and validity fact-checking.
Based on a recommended proposal located at: http://strategy.wikimedia.org/wiki/Proposal:Develop_systems_for_accuracy_review
Identity
Name: Ryan Miller
Email: millerrv@purdue.edu
Project title: Peer voting-based bias indicator for controversial topics
Contact/Working Info
Timezone: GMT-6 Central Standard Time
Typical working hours: 10AM - 10PM
IRC or IM networks/handle(s): AIM: rvmiller89
Project Summary
Fact-checking has become an important issue when determining the veracity of Wikipedia articles. Internal sourcing is the traditional and current approach to establishing reputability on Wikipedia, but an electronic and Internet-based medium calls for non-traditional techniques.
I propose an interactive voting system in which Wikipedia users can rate the potential bias in an article in a secure and controlled fashion. Once statistically analyzed for proper distribution and sample size, this indicator will provide insight into the controversial nature of articles by flagging articles deemed too biased to fit Wikipedia's guidelines. This system can reduce disputes over subject matter, and can replace the time-consuming and sometimes unreliable method of the Third Opinion.
Voting security will be an important aspect of this system in order to prevent directed attacks or rigging by nefarious groups. One method could grant only logged-in Wikipedia members the ability to vote, and could more heavily weight the opinion of those members with more active Wikipedia participation and a history of unbiased article generation.
About Me
I am a sophomore majoring in computer science and minoring in psychology at Purdue University in West Lafayette, IN. I have taken coursework on programming in Java and C, data structures, algorithms, operating systems and computer architecture. I am an avid proponent of Wikipedia and its usefulness as a collective resource of knowledge. I authored an article on the U.S. court case Dean v. Utica after researching the topic for a class project and finding no Wikipedia article. My interests include software and hardware, journalism, reading, writing, vintage Apple, and motor scooters.
Deliverables
Required Deliverables
- Research current guidelines on Wikipedia unbiased writing (ie., the Neutrality Policy)
- Design on-page voting system to allow users to vote in an unobtrusive manner (Perhaps using SecurePoll WikiMedia extension)
- Determine correct sample size needed to deem ratings as statistically significant (Possible approach - Lower bound of Wilson score confidence interval for a Bernoulli parameter)
- Prevent attacks on voting systems (determine reliability of SecurePoll extension)
- Train voting systems to place emphasis on more established members who display: 1) active participation on Wikipedia and 2) a history of unbiased article generation
If time permits
- Determination of some controversial articles in need of such a voting system
- Refinement on the voting weight algorithm
- Evaluate feasibility of automatic flagging of articles passing a significant threshold of bias (research topic only)
- Graphical representation of results
Future research
- Natural language processing to detect levels of neutrality in articles and reliability of sources
Project Schedule
Weeks 0-2 - Research current guidelines on Wikipedia unbiased writing and design on-page voting system to allow users to vote in an unobtrusive manner
Weeks 2-4 - Continue implementing voting system and determine correct sample size needed to deem ratings as statistically significant
Weeks 6-8 - Research ways to prevent attacks on voting system and test for vulnerabilities. Create method for training voting systems to place emphasis on more established members who display: 1) active participation on Wikipedia and 2) a history of unbiased article generation
Weeks 8-10 - Finalize voting system and study an automatic flagging system for articles deemed too biased
Participation
I have an open summer and a willingness to work every day. I can report progress regularly and chat often.
Past Open Source Experience
This is my first summer participating in Google Summer of Code, so I haven't yet worked on any open source projects. I have experience with Linux, and will be writing a system call from scratch as part of my operating systems class this semester. As part of an internship last year with a startup computer security company, I used the Lightweight Directory Access Protocol (LDAP) and Apache tools to create a directory-searching component for their software.
Any Other Info
My internship last summer focused on computer auditing and data mining, so I have experience searching for data and processing the results. Also, as part of my data structures course, I worked with a partner to write a web crawler and search engine in C++ that emulated the basic search features of Google and efficiently stored results in AVL trees, hash tables, and binary search trees. Feel free to e-mail me or contact me on AIM. I'm also on Facebook under my name and location. My resume is available at http://rvmiller.com.
References
http://en.wikipedia.org/wiki/Wikipedia:Third_opinion
http://en.wikipedia.org/wiki/Wikipedia:NPOV
http://www.mediawiki.org/wiki/Extension:SecurePoll
http://www.evanmiller.org/how-not-to-sort-by-average-rating.html
http://strategy.wikimedia.org/wiki/Proposal:Develop_systems_for_accuracy_review
Community Discussion
Do you have a thought about this proposal? A suggestion? Discuss this proposal by going to Proposal talk:Peer voting-based bias indicator for controversial topics.
Want to work on this proposal?
- .. Sign your name here!