3 ms·

Wikipedia is chock full of human bots that try to turn it into a relational database or some kind of graph of human knowledge. Just look at all the categories H

Wikipedia is chock full of human bots that try to turn it into a relational database or some kind of graph of human knowledge. Just look at all the categories Hitler is in: - Suicides by firearm in Germany - German founders of automobile manufacturers - Austrian anti-communists - ... Of course Wikipedia (or rather MediaWiki) is completely terrible for this purpose. It's a completely unstructed raw text storage. You can find date A cited on one page for event X, then go to the page on event X and find a completely different date. It's all just text, and the human bots are overwhelmed with keeping it synchronized. Of course the human bots will be angry if you spit on them in some dead tree op-ed thinking their valiant efforts toward reducing Wikipedias text to ever smaller isolated atoms are hurting some political ideal. (Of course the people over at Wikipedia have recognized that the problem is insurmountable the way they are approaching it. They have created WikiData in response: https://www.wikidata.org/wiki/Wikidata:Main_Page https://www.wikidata.org/wiki/Wikidata:Main_Page)

Svip 13y ago
As someone who maintains the largest Futurama wiki (albeit not as active anymore, but I still host it), I can recognise the problems that Wikipedia is faced with. Of course, at a much much smaller scale, as we only have 3,000+ content pages (which is a lot for a Futurama wiki). And we have had to deal with the usefulness of indexing articles appropriately. That is; giving them categories (and whether new categories is going to help anyone). But since our wiki is so much smaller, we don't use bots but do it by hand (although, we do have one bot to do some basic cleaning up, but nothing truly intelligent). And the issue isn't just with Wikipedia or our wiki; it's practically with every wiki (even non-MediaWiki wikis). This has led to the creation of Semantic MediaWiki, an extension that is supposed to maintain more rational relationships of data and make it easier indexable. But most wikis I've seen using Semantic MediaWiki, doesn't really look like a content site, but rather a data site. Which is nice for the purpose, but it makes you wonder if MediaWiki really was the right choice to begin with.

Svip 13y ago

As someone who maintains the largest Futurama wiki (albeit not as active anymore, but I still host it), I can recognise the problems that Wikipedia is faced with. Of course, at a much much smaller scale, as we only have 3,000+ content pages (which is a lot for a Futurama wiki). And we have had to deal with the usefulness of indexing articles appropriately. That is; giving them categories (and whether new categories is going to help anyone). But since our wiki is so much smaller, we don't use bots but do it by hand (although, we do have one bot to do some basic cleaning up, but nothing truly intelligent). And the issue isn't just with Wikipedia or our wiki; it's practically with every wiki (even non-MediaWiki wikis). This has led to the creation of Semantic MediaWiki, an extension that is supposed to maintain more rational relationships of data and make it easier indexable. But most wikis I've seen using Semantic MediaWiki, doesn't really look like a content site, but rather a data site. Which is nice for the purpose, but it makes you wonder if MediaWiki really was the right choice to begin with.