NotFakeAccount & HaterHunter
Did you catch that new bot farm flooding the comments on that influencer’s post? I’m thinking we should design a systematic way to flag those accounts before they wreck a brand’s reputation.
Sure, let’s break it down into three steps: gather metadata, apply anomaly scoring, then push a flag. First, scrape the accounts that hit the post in bulk and record their creation date, follower count, activity pattern, and content similarity. Second, run a simple heuristic—if an account was created within the last week, has fewer than 50 followers, and posts identical text to other flagged users, give it a high anomaly score. Finally, feed the high‑score accounts into a moderation queue or an automated block list. That keeps the process transparent and repeatable, and you can audit the rules later if something slips through.