Tuesday, August 17, 2010

Why I left Wikipedia

An article in this week's Newsweek reports that Wikipedia has been floundering since the spring: "Thousands of volunteer editors, the loyal Wikipedians who actually write, fact-check, and update all those articles, logged off-- many for good." The WSJ first reported the fallout almost a year ago when it was discovered that 49,000 English editors left Wikipedia during the first three months of 2009 compared to a loss of 4,900 during the same period in 2008.

Update: As one of the comments below states, the WSJ article was hasty in their conclusions. It all hinges on what you call an "editor", and a more balanced definition suggests that editors are not leaving Wikipedia in droves.

As the Newsweek article points out, there are a number of reasons why Wikipedia may be stagnating. There are so many articles already present that there is little new ground to break. Some may be scared away or frustrated by overly aggressive editors. Or perhaps "most people simply don't want to work for free."

Some research at Georgia Tech shows that editing a Wikipedia article is very challenging for computing newbies; the "Editing this way will cause your IP address to be recorded publicly" message causes lots of confusion, and this certainly prevents many from joining the ranks of Wikipedia editors.

I have always been a Wikipedia fan. I first started making serious contributions in 2004 when I was beginning my PhD research and discovered that many of the new concepts I was being introduced to simply didn't exist in Wikipedia.

I wrote a number of articles from scratch like web archiving, web search query, adversarial information retrieval, and URL normalization and made a significant number of edits on other technical topics. I was motivated in part by being the first to write the articles and the fact that I would likely refer back to them as reference material as I continued my research.

However, I found that keeping vandalism at bay and fighting poor edits was quite time-consuming. Some articles that I valued quite highly like web crawler needed tons of work, and although the desire was there, I just didn't have the time... I was trying to complete my PhD, and maintaining Wikipedia articles was not paying the bills.

I had an ah-ha moment at a conference a few years ago when someone quoted from Wikipedia's article on digital preservation, and I could have sworn I had been the sole author of the quoted piece. Wikipedia was given credit as the source, not me. That didn't bother me all that much, but it did make me realize that contributing to Wikipedia is often not in the interests of academics who are often judged by the amount of citable material they produce. Someone citing what you wrote in Wikipedia doesn't "count" like someone citing what you wrote in a journal article.

Over the past year or so, I just have lacked the motivation necessary to put time into an anonymous forum. My time is expensive, and Wikipedia is not paying. It's hard enough just to find time to edit my blog!

I still think Wikipedia is extremely valuable, and I hope it never goes away. I regularly send my students there and encourage them to make a serious contribution.

Have you seen The Book of Eli? At the end of the movie, a group of people are attempting to restore some of the greatest literary works of mankind. They are quite happy to have nearly a complete set of Britannica encyclopedias. No mention is made about the remnants of Wikipedia. :-(

7 comments:

  1. Unfortunately they probably didn't have the web in the Book of Eli, they didn't even have electricity!

    ReplyDelete
  2. You're right... no electricity and no Web means no Wikipedia. Maybe someone needs to start bounding volumes of Wikipedia just in case. ;-)

    ReplyDelete
  3. The Wall Street Journal story was complete garbage. The measurements that were used to calculate the "loss" of editors is time-sensitive, and the person who compiled them failed to do a sensitivity analysis—one which reveals that based on the measures used, one could "reveal" a loss of editors at just about any point in time. Eric Zachte pointed this out on his blog at http://infodisiac.com/blog/2009/12/new-editors-are-joining-english-wikipedia-in-droves/

    I have to sympathize with your lack of time, though. It's hard to balance doing the good things we'd like to do with the practical things we must do.

    ReplyDelete
  4. Thanks for your comment, Nihiltres. I've updated the blog post based on this new information.

    ReplyDelete
  5. I think there would be a resurgence in contribution if Wikipedia started including conventions from Semantic MediaWiki. The whole list curation problem in Wikipedia wouldn't exist anymore! The only problem is that it would complicate wikitext even more.

    ReplyDelete
  6. I hadn't heard of SMW until you mentioned it... looks like it would be a very valuable tool if a typical user could add semantic notations in a simple manner. It would be even better if Wikipedia could automatically convert simple edits into semantic constructs. :-)

    ReplyDelete
  7. That is what DBpedia is doing. They take all the structured content that appears in the right hand side boxes on a wikipedia page and they put it in their own triple store which is accessible by anyone. But yeah, it would be better if wikipedia automatically made that information available through a SPARQL endpoint.

    http://dbpedia.org/About

    ReplyDelete