I've been working on this for almost a month but first I'm trying to scrape users based from different countries by their local board posts, by simply going to a local board and scrape every comments in every thread and in every page (this is really hard). I hope I could patent this idea so my work and time spent studying scraping wouldn't be wasted. After that, I think it would be easy if I had stats with the links of local users and simply scrape when they are active (but would take me another month studying another prog language I guess).
I appreciated your time to learn and your plan to build up a project. Honestly, I don't know how to scrape website's data.
Quality is better than quantity.
Figures on active or inactive members do not reflect the real situation on the forum, such as the average quality of posts or threads created by members recently. Imagine that an active local board can be detected by higher number of posts in latest few weeks compare to the other local boards. But such figures can not make sure the quality of that active local board that is probably a spam place. More active more spam endemic inside.
A wider view, the forum used to be more actively in the past, late of 2017 and early of 2018, but now we have a very better forum (less spam, higher quality).