Google Knol
Topics: Thoughts | July 23, 2008 @ 5:46 PM
I just read about Google’s new Knol. Basically, it’s Google’s take on online encyclopaedias. We all know Wikipedia, and while I’ve defended it in the past (and currently maintain my position) it does have some undeniable issues:
- Verifiability. Wikipedia is great to get some information or learn some new things. But it just cannot work well as a primary source, for instance, because the quality of the articles is often less-than-perfect, and is itself often lacking in proper citations[citation needed]. If I remember correctly, the editors themselves only approved a couple thousand articles to be considered citation-worthy – and there are several million articles on Wikipedia. Wikipedia is brilliant in the general availability of knowledge that it provides, but it’s unviable as a citation source in papers or even elementary school projects (that is, if we want to teach our kids what verifiability means).
- Lack of objectivity is another problem that can plague Wikipedia. While its standards (and those that enforce the standards) do what they can to prevent and fix this problem, giving everyone anonymous write access means that it’s easy for there to be “wars” on “controversial” topics. For instance, it’s perfectly valid (and, I believe, important) to discuss Intelligent Design in its social and historical context, but since it isn’t science it shouldn’t even be mentioned on scientific articles. But there are mounds of people who think the exact opposite, and would gladly try to add it in there. Now, it’s simple enough to change it back, but then someone just needs to change it again to ruin it.
- Authors! Like I’ve said before, anonymous access enables little bits and pieces of information to end up on Wikipedia, which is great – but if someone who is a known expert says something, wouldn’t you believe them over some random boob? Wikipedia provides no easy way to see how reliable an article is from that perspective.
Google tries to address some of these problems. For one, Knols are specifically attributed to authors, and only the author can change things on that page (viewers can make suggestions). Of course, multiple authors can write on a single subject, but what they write goes into separate pages. This, I believe, addresses the three above problems. If you have an author’s name, you can specifically cite that author. And while it doesn’t directly address lack of objectivity, the fact that a Knol’s author is an essential part of that Knol keeps the fact that there is a specific person behind the article and their opinions are likely to colour it (kind of like a blog). These authors can already do things like this, but Knol provides a centralized location to which everyone has (read) access. And, of course, random boobs can write articles, but no one’s going to listen to them because we know it’s just Rando Mdick writing the article, not an authority. And since the articles are ranked by popularity, similar to how Google ranks websites, these random boob articles are probably not going to show up to most people too often.
I do see some problems though:
- Impersonation. Now, I’m sure that there are some measures in place to guard against this, but it still seems perfectly possible.
- The double-edged sword of popularity-ranking means that a large quantity of crazies could bring certain articles up to the top (like Google bombing). Since we all know that stupid people vastly outnumber everyone else, this could be a problem in that certain articles (e.g. Intelligent Design as science) could be bumped up. On the other hand, though, Google’s system works well enough (better than other search engines!) for websites, so hopefully this will carry over.
1,116 views | Trackback
