Research:Using information sources to drive contributions to Wikipedia
We wish to study how contributions to online collaborative work can be enhanced by supplying relevant source content. Two challenges which member-maintained online communities such as Wikipedia may face include:
- Acquiring new content
- Acquiring verified content (content with sources)
In response to these challenges, we have investigated news feeds and their potential application to the process of editing Wikipedia. We are developing a tool which integrates Twitter and RSS, allowing users to input their preferred news sources and receive recommendations for possible contributions to Wikipedia based on their news sources. We hope that this will make it easier for editors to put new content with citations into Wikipedia. To test our tool, we will invite Wikipedia editors to participate in a user study and monitor their editing activity facilitated by our tool. We will analyze this data along with participants’ past editing behavior to assess whether the quality, quantity, and diversity of Wikipedia contributions by study participants are affected.
We chose our invitees sampling from three pools of Wikipedia contributors:
- All contributors registered on Wikipedia (randomly)
- Contributors who have made an edit in the last 30 days on Wikipedia (randomly)
- Contributors with very recent activity (sequentially from Recent Changes at the time the sample was taken)
We are committed to recruiting participants without being annoying. Once we begin recruitment, we will post an invitation to join our tool on the talk pages of the selected Wikipedia editors. Those who do not respond will not be contacted again. Editors who wish to participate will fill out a consent form before creating a user account in our tool. Additionally, one participant, selected at random, will receive a book token worth fifty US dollars.
Here is the recruitment message we intend to send. We will split the effort of recruiting amongst our team; hence, the bracketed researcher name.
- Hello [username],
- I'm part of a team that is conducting a research study investigating how to help Wikipedia editors find interesting content to edit. More specifically, we are interested in seeing if content from recent sources of news can be used to enhance Wikipedia editing. We have created a tool, called WikiFeed, that allows you to specify Twitter and/or RSS feeds with news sources that are interesting to you. WikiFeed then helps you make connections between those feeds and Wikipedia articles. We believe that using this tool may be a lot of fun, and may help you come up with some ideas on how to contribute to Wikipedia in ways that interest you that you, and for which you are qualified, that you had not previously considered. Please participate! To do so, please visit [url to be determined] and sign up. If you have any questions, please feel free to ask via my talk page, or by email at [email address].
- Thank you, [researcher]
Benefits for WikipediaEdit
We hope that our tool will ultimately benefit Wikipedia by enhancing contributions in terms of quality, quantity, diversity, and credibility. By giving editors news they’re interested in and, in turn, making recommendations for Wikipedia editing based on those news sources, we hope to make editing easier, more effective, and more enjoyable for the Wikipedia community. We would love to see Wikipedia - as both a community and an encyclopedia - continue to thrive, and we hope our tool will only continue to make Wikipedia stronger.
Sharing our FindingsEdit
We intend to present the results of our work at a research conference such as IUI or WikiSym, and publish the results in an associated paper. If the online tool that we create is successful, we would make it available to the Wikipedia community at large.
Wikipedia Policies, Ethics, and Human Subjects ProtectionEdit
- Participant recruitment starting July 2
- Tool Development until July 16
- User study begins July 23
We would love to hear your input! Any questions or comments can be directed towards David Musicant.