User:Mumia-w-18/Article Half Life

Please place any responses in the "Responses" section at the bottom. Thanks.--Mumia-w-18 09:19, 10 November 2007 (UTC)

Article Half Life

Nature of the problem

Undefended articles on Wikipedia have half-lives--in other words, they will decay due to vandalism over time. There are 100-300 vandalism events per-hour on Wikipedia. Articles that are zealously defended by their interested editors can survive in this environment, but articles that are not being watched have no chance.

And it seems no one is watching the bot-created artitcles.

I'm going to assume, for the sake of argument, that the article half-life is three years. In three years, half of the 10,000 bot-created articles will have been vandalized. In another three years, half of the remaining "good" articles would have been vandalized--resulting in 75% of the articles being bad.

I read the pages that AndrewGNF gave me (here and here). No where did I see anyone claim that they were going to try to defend these bot-created articles. Evidently people in the Biology Project want these articles, but they are not willing to defend them.

I don't have to remind you that Wikipedia is not a normal encyclopedia, and it's definitely not static--the Biology Project is going to have to fight for these articles, because you're the only ones who can.

Possible solution

I wish I'd been around in May 2007 to warn people of what a bad, bad idea this would be. It's too late now; that ship has sailed. The members Biology Project are now going to have to decide how much they want these articles. A single editor can probably do a reasonably good job of defending 100 articles, so I suggest that the Biology Project assign 100 articles to 100 members. These people will have to be willing to defend these articles over the next fifteen years (or however long it takes for the Wikimedia Foundation to decide to lock down the data, and when it is locked down, Biology Project members will be called upon to perform a full audit of all 10,000 articles).

Data this technical should not be editable by the public. Doing this on Wikipedia is a bad, bad idea. But the articles are here, and more are coming; Biology Project members are going to have to get on their boxing gloves and fight for their articles.--Mumia-w-18 09:19, 10 November 2007 (UTC)

Responses

The concerns apply to the whole Wikipedia project... You can't force editors to patrol the articles they create, and without something like stable versions, the heap of neglected articles will be always a problem (see here for an example I discovered using User:Gurch/Test). In my experience the articles in the WP:MCB project are on average mantained in good shape, thanks to a strong community and dedicated users (like Arcadian). In my watchlist, most nonsense or malicious edits hit well known articles like ribosome and protein: in some days these articles alone account for almost all bad-faith edits I see in molecular biology articles. On the other hand, technical articles like Bone sialoprotein receive virtually no vandalism, and even anon editors often add useful content; furthermore, the edit rate is so low (few edits/year) that one can easily patrol tens or hundreds of such articles.

A bot-generated group is also far easier to check. This spring, 500+ RNA families were imported using a bot, and there is a "collective watchlist" (User:Jennifer Rfm/Rfam WatchList) that can be checked using the special page "Related changes". In the last month, you can see only 1 vandalism edit (reverted by a bot) on 500 articles, and many good edits.--Banus 12:53, 10 November 2007 (UTC)

If I was wrong and people are watching these pages, then the project might work out as expected. Stable versions seems like an interesting idea also.--Mumia-w-18 14:27, 10 November 2007 (UTC)

Vandalism rates decay exponentially as you increase the technicality of an article. Protein and organelle get a lot of vandalism, while some of my more arcane stubs like Ellman's reagent or trypanothione have never been vandalised. Tim Vickers 18:07, 10 November 2007 (UTC)

If I could make a more "personal" argument:

I'm a PhD student that studies FOXP2. I had little experience with Wikipedia, but a couple of months ago my boyfriend asked me if Wikipedia had a FOXP2 article. It did, but it wasn’t a very good article. My boyfriend prodded me to edit the article. I finally did, and now I’m hooked. I’ve edited many other articles, have 400+ molecular biology-related articles on my watchlist that I revert vandalism on, and contribute to the Molecular and Cellular Biology Project. If there hadn’t already been a FOXP2 article on Wikipedia, I never would have taken the initiative to learn how to create an article, how to format references, and how to interact with the Wikipedia community.

As for vandalism, I agree with User:TimVickers and User:Banus. Although I have 400+ articles in my watchlist, if you look at the recent changes you would think I had only 20+: Chromosome, Genetic engineering, DNA, Evolution, Enzyme,….

Having articles, even if meager stubs, will draw scientists in. Like myself, scientists may initially contribute to an article they feel they are an expert in, but may not want to create such an article from scratch. Without scientists, molecular biology articles won’t be breathing, vibrant articles that show the current views in the field. And most scientists that I know want the world to understand why their protein of interest is important, why it’s exciting, and how understanding it is helping humanity. They want the public to have the information to make informed decisions about public policy and their lives. Although this isn’t an argument about Wikipedia policies, I feel it is in line with the spirit of Wikipedia. I believe that the ProteinBoxBot is creating an enormous amount of good, and that is why I am contributing to this project. Forluvoft 22:50, 10 November 2007 (UTC)

That's a pretty good vote for having the article-stubs. It's true that we don't know how many people will participate in improving these articles. And if the technical nature of these stubs allows them to fly "under the vandals' radars," then the project should work pretty much as expected.--Mumia-w-18 23:35, 10 November 2007 (UTC)