A stackoverflow discussion about how to use python to make an xmlhttprequest. I looked this up because that's how you're supposed to interact with unmark's pseudo API. The accepted answer shows how to do it with scrapy, but I think it could be adapted for use with the requests module. The HTTP method would be POST, and one of the headers would be "X-Requested-With: XMLHttpRequest". It's worth a shot, at any rate.
The wikipedia page about how XMLHttpRequest works. This seems like a pretty straightforward description of the standard.