3 ms·
Text Mining Wikipedia for Misspelled Words
- mukyu 15y agoOdd that they didn't use the API or action=raw to get the wikitext and instead just scraped and went digging. It would require post-processing anyways (rip out templates, wikilinks, reflists, whatever) but it would be easier to isolate what is actually in the page.