Metaculus / Metaculus/metaculus

Comment Spam Detection

Open
#1,825 1 comment 0 reactions 2 assignees View on GitHub

@CodexVeritas is already working on this.

Since Dec 26, 2024.

Low Priority Stale Technical
Dominant language
TypeScript
Stars
113
Forks
34
Avg merge
16h 34m
Merged PRs (30d)
30

Description

This is to track detecting comment spam.

  • Phase 1:
    Some comments are of the form:
    "@[ random_fake_username ] ( www.spam_link.com ) Something that is just a completely generic but on-topic comment that is likely bot generated and probably can't be easily detected".
    What they're doing is making a comment that looks like it's tagging a real user, but it's just a spam link.
    This is super easily detected by a simple algorithm, so doesn't even require a call to GPT. I think this would be a very quick first pass solution.

  • Phase 2: (tags are currently for this phase)
    use GPT to look at the raw text of the comment and see if there are any suspicious links. If there are, flag user as spam.

  • Phase 3:
    actually parse the whole comment and block if it's obviously spam, but I've seen so few comments like that that I'm not currently worried.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.