it's mostly a matter of pattern-recognition...

in this case, I identified what was happening: an existing comment on a post being duplicated. then, I identified that a few words had been changed...

then, as I looked at more of the users' comments, I saw the exact same thing happening over and over.

that'd be alot of effort for an individual to go through. and, given that the user claims to be based in Bangladesh, I could infer that their English is probably nowhere good enough to actually have changed up the few words they did themselves.

another giveaway was that one of the comments going out from their post was in Korean. not likely they also read/write fluent Korean. and if you copy-and-paste that Korean in the "find" option of your web browser, you'll find it's actually just duplicated content from the post itself!(and, their profile also says they have an interest in AI development ).

add those points all up together... you can be pretty sure this was a bot.

for other situations...

mainly, finding a pattern of repetitive posting - whether it's outright spam, or just follows a very clear pattern with some random fluctuations that would seem very unusual for an actual person to write themselves...