Today, I've found in the Wikinews logfile two additional example for a quickly deleted article. I was fast enough to take a look into the content before it was deleted. The first article was a single paragraph without a source, while the second article was a longer text of around 5 paragraph and didn't have a source too. Both articles were deleted in around 30 minutes after creation and this kind of admin intervention was predictable. The most basic simple rule for understanding Wikinews is, that an article gets deleted by 100% if no source is given at the end.
Now i'd like to improve the prediction model a bit to analyze under which conditions an article gets not deleted but is put into the normal review queue. A recent example has the title “Newly discovered theropod shows another avenue to flight, say scientists”. After creating the article and 30 minutes later no admin intervention was there instead the author had made additional edits and two days later the article is still in the system so it seems, that the admin is positive for the article. Let us take a look into the initial version which was submitted to Wikinews.
The article contains of around 2 paragraphs about a dinosaur. No images are in the article and at the end two sources are given. Both sources can be found in Google news as well. That means after copy&paste the headline from the wikinews article to Google news the source is found. It seems that the combination of two textual paragraphs plus 2 Google news sources are enough to overjump the entry barrier and get not deleted by the admin.
But, even this rule seems to be good prediction rule, it is unclear if such an article creation can be repeated by another person. The question is: what will happen if a new user not known before puts creates an article which contains of two Google news sources and two paragraphs. Will the article remains in the system or gets he deleted like all the other articles? I don't know. In this point it's hard to predict the Wikinews project. It might be possible that they are open to everyone, but it might also be possible that only certain users can submit new content.
Let us summarize all the known situation from the past.
Case1: article without a source -> gets deleted by 100% in under 30 minutes
case2: article with a source of encyclopedia dramatica and a single sentence -> gets deleted in under 30 minutes by 100%
case3: longer article with a picture and a single source from a medium blogpost -> gets deleted in under 30 minutes by 100%
case4: two paragraph article without a picture and two sources from Google news -> remains in the system. It is unclear of which reason exactly.
No comments:
Post a Comment