We occasionally get spam in the suggested edits like the following:
- Not a single word that can be found in the English dictionary (except perhaps URL).
- Identical text in the suggested edit and edit reason.
- Wordpress style links such as
- suggested anonymously
It seems to me that the modus operandi of this spammer is clear enough that their suggestions can easily (and accurately) be rejected by the SE engine using simple heuristics based on the observations above.
Another possible idea (in addition to the above) would be to reject any suggested edit that drastically lowers the post's "quality score", regardless of whether it is spam or not. However, we need to fix another problem before this is implemented — SE's low quality filters do not consider the above spam "low quality"!
This came as a complete shock to me. I posted the above spam as an anonymous answer here to grab a screenshot of the automatic community flag for "low quality" (which is raised as soon as the answer is posted) or the "this question does not meet our quality standards" message. My goal was to try and argue that if the above gibberish is rejected by the quality filter, then it should be rejected by the suggested edit filter too, but surprisingly, the above answer sailed through just fine (but this didn't! o_O)
It would be helpful if the developers added spam checking in the suggested edits and also tweaked the LQ filter to not allow answers like the above to pass through. While it is a no-brainer to reject these suggested edits, it still is a waste of time doing something that can and should be handled automatically. Besides, I'm sure this spammer is a nuisance all across the SE network, so implementing this will benefit more than just Mathematica.SE.