New poster, ella1985, joined today, only three posts so far, looks like an AI in every case.[
](http://boards.straightdope.com/sdmb/showthread.php?p=13964689#post13964689)[
](http://boards.straightdope.com/sdmb/showthread.php?p=13964687#post13964687)[
](http://boards.straightdope.com/sdmb/showthread.php?p=13964686#post13964686)The misspellings are a nice touch.
So the plan is to trick us into believing it’s human, and then it’ll bust out the links for unniveersity d1ploomas and peen1s enl4rgment and whatnot?
Yes, definitely a bot. I’ve banished it to the land of wind and ghosts, so the posts are now deleted.
You sent them to Bodie, California?
http://www.visitcalifornia.com/Must-Sees/Desert-Wind-and-Ghosts-at-Bodie-State-Historic-Park/
I would like to understand this better.
Is there some server somewhere that creates an account on a message board, randomly picks threads, and posts badly misspelled generic and vague replies?
For what purpose?
None of the quotes above seem like a plug for some other website…
I would say those posts were cover and the spam would have come later. It’s just supposed to look good enough to pass a cursory inspection and not get deleted, I think. It doesn’t have to look like a trenchant insight.
Not that I know much about this sort of thing, but I do clean up a lot of botcoments on a blog at work – maybe bots like **ella **just don’t competently distinguish between boards that do and do not allow links as part of usernames or sigs. I delete lots of poorly worded but polite and kinda-sorta-on-topic comments whose only spamliness is in the userprofile.
The third post certainly had no content, and the first one, I could see being just “<subject of thread> is a problem…”, but I’m wondering about the auto-parsing that would have come up with the second post. Did it pick out “nothing” as the relevant word in the OP, or just notice that it was a word that was being repeated a lot in other posts?
And my guess is that the misspellings are to make it harder to recognize repetition of posts. If a bot used the exact same text for something like that third post in multiple threads or on multiple boards, it could be caught that way, but by putting in a random set of “typos”, each post can be made different.
I think ella1985 make interesting sepculation…of course, theirs two sides always isn’t their?
The second quote, I guess a decently written bot might have been able to link “definition” and “nothing”. I guess.
The third quote is extremely generic, could be plugged into almost any thread here.
How do you determine this? Do you check the IP address against other sites that have already determined it is a bot?
Easy. You ask it if it would like a delicious bowl of hex nuts and a tall glass of machine oil to wash them down with, and if it says “AFFIRMATIVE”…
Assuming it’s a bot, the programmer has no opinion on regionalism. It would just be programmed to detect a keyword and say that it’s a good or bad idea so it looks like it’s leaving a comment.
In theory, it’s so we don’t notice the spam. It’s more subtle than inserting a bunch of links to a bootleg shoe site or amateur porn, but only a little
That’s the easiest technique to use, yes.
I’m not seeing any of her posts in those threads–or any of her posts at all. Has she been disappeared into the corn field?
Yes. Once I saw this thread and confirmed she was a bot, I banned her and deleted her posts.
Here’s the thread in question.
The bot took the entire subject line of the thread and dropped it in the sentence:
But, pulling a thread title from Cafe Society, you could have had this:
Just want to mention taht you developed avery intresting subject here! Mom Brady got the crabs from John Lindsay is a problem and people have to be aware of it.
Next step: color it purple!
It obviously has multiple different templates to choose from for crafting replies. Maybe it only uses that one for thread titles below some length.
Is John Lindsay that captain on Deadliest Catch? Them is big crabs.