In the recent months, we’ve been getting more blogspam accounts, and the administrators have been discussing behind the scenes on how to deal with it. Blogspam is against the rules of this Lemmy instance and is treated the same as any other spam. That is, offending posts will be removed and blogspammer banned. I thought I’d share my thought process of moderating stuff like this.

Blogspam is kind of a controversial topic and has a lot of grey areas. It basically involves accounts seemingly made specifically to post links to a specific website, usually with the intent of generating ad revenue. Herein lies the grey area, because simply posting links to your own website or a website you like isn’t spam, nor is it against the rules to post websites that have ads, nor is it against the rules for an organization to have an official account on Lemmy, so it becomes a problem of where to draw the line. You can also run into problems where it’s hard to tell if someone is intentionally spamming or if they’re just enthusiastic about the content of a site.

That said, here are my general criteria on what is considered blogspam, with some wiggle room on a case by case basis:

  • Does the user only post links to one or a few sites? Do they have any other activity, such as commenting or moderating communities?

  • How often does the user post? For example, it might not be reasonable to consider an account to be blogspamming if they only post a few articles a month, even if they only post one site.

  • Does the user post the same link repeatedly? Do they post to communities where it would be off topic?

  • Is the user trying to game the search feature in Lemmy by including a large number of keywords in their title or post body?

  • Are the links posted “clickbait” or otherwise designed to mislead the reader?

  • Is the site trying to extract data or payment from readers? Examples include forcing users to sign up or pay for a membership before letting them read the article.

  • Is the site itself well-known and reputable or obscure and suspicious?

  • Does the site have an “inordinate” number of ads? Are the ads intrusive (autoplaying video ads versus simple sponsor mentions for example).

  • Is there evidence that the user is somehow affiliated with the site? Examples include sponsored links or having the username be the same as the site name.

  • Is there evidence that the user is a bot?

Not all of these have to be satisfied for it to be blogspam, and it’s usually up to the administrators to make a rational decision on whether to intervene.

Note that these criteria apply to sites that are generally benign, but is being posted in a way that might count as spam. If the site contains malware, engages in phishing, is blatantly “fake news”, is generally malicious, etc, those alone are reason enough for it to be removed, and would constitute as a much more serious violation of our rules.

I’m open to feedback on this, feel free to discuss in the comments!

  • @GrassrootsReview@lemmy.ml
    link
    fedilink
    13 years ago

    Is the main side-wide problem that these posts turn up in the search results? Otherwise they are only seen by people subscribing the moderated communities people subscribed to?

    If yes, an option for (low volume) spammers could be to only exclude them from search results and otherwise have the moderators of the communities and their downvoters deal with them. As you already write many cases are grey areas, so maybe in such cases such more subtle mechanisms are enough.

    As only removing from the search results is less disruptive one could moderate more posts this way.