OpenStreetMap

Can we please fix the SEO Blacklink Spam?

BlackHat SEO

I have been mapping here for a few weeks now and what caught my eye from day one was the pretty large amount of spam here and how long it sometimes takes until it gets removed - IF it gets removed at all…

And while diary entries usually get removed pretty quickly (which is still annoying as they show up in the RSS feeds and on Social Media), some accounts with spam in the user descriptions are over 1 year old as I just found after reading this entry.

So What’s the Solution?

User Descriptions:

  • Someone has to work his way trough the list in the Wiki and remove the profiles.

  • Maybe implement a report Button

User Diaries:

  • Only publish them after a moderator has approved them.

  • Require accounts to have a certain age or amount of edits (the 2nd one could obviously backfire as it might lead to vandalism)


With all the active people and programmers involved in this project there must be better way to deal with this and get it under control.

Don’t you think? Anyone got any better ideas?

Discussion

Comment from TomH on 9 March 2014 at 13:10

It’s just not a major problem - the vast majority is blocked as soon as it is entered and the small amount that gets through is blocked by me as soon as somebody makes we aware.

In the case of english language spam I will normally see it in my RSS feed within an hour or two - other languages are more problematic.

User spam is less likely to be spotted, but it also isn’t a major issue because nobody will actually see a user page for a user that doesn’t do anything unless they actually go hunting for it for some reason - it’s not like the diaries which will appear in people’s RSS feeds.

It is probably reaching the point where we need to build a reporting system, but that’s quite a lot of work to get all the necessary backend bits in place and nobody has had time yet.

Comment from RobJN on 9 March 2014 at 17:38

It would be nice to see the user diaries and community blogs get used a bit more. I think user “mcld” is working on updates to the blogs.openstreetmap.org page at the moment. I’m hoping that will integrate it back into the new osm.org design.

But… we could do with a few more active programmers. Fancy having a go yourself? :-)

@TomH: Do we need to look for some new moderators (particularly non-English language)?

Comment from Wynndale on 9 March 2014 at 18:39

As we get an amount of spam in Polish is it worth recruiting a Polish speaker?

Sometimes spammers post platitudes as diary comments trying to link back.

Comment from vvoovv on 9 March 2014 at 19:18

So for now, the solution for a non-English language is to send a private message to TomH . Right?

Comment from AndiG88 on 9 March 2014 at 23:16

Sometimes spammers post platitudes as diary comments trying to link back.

It’s pretty much the case 95% of the time, which is why it should no be that complicated to detect the spam even if it is in a different language. I mean there is really no good reason why a diary should link to some polish furniture store and if in doubt you can still check the profile it usually always has zero map edits.

Comment from TomH on 10 March 2014 at 12:11

@RobJN Not sure how more moderators is supposed to help? Firstly because currently only full admins can do anything about this and secondly because the problem is mostly about detecting the spam rather than removing it.

@AndiG88 If automatically detecting it’s so easy then I look forward to seeing the code, as I’m sure does the author of every blogging platform in the world.

Comment from AndiG88 on 10 March 2014 at 12:21

the problem is mostly about detecting the spam rather than removing it.

There is a whole Wiki Page full with links to Spam Profiles.

If automatically detecting it’s so easy then I look forward to seeing the code, as I’m sure does the author of every blogging platform in the world.

Well, I think one automated mechanism that would work (if it is possible to code) is that you need a certain amount of edits to publish a diary and otherwise you can’t or it will need mod approval. (Maybe even require that for links in user descriptions, so it just displays the text for new accounts).

Obviously you can’t automoderate links, my point was just that you usually don’t need to understand the language to see if a link is spam or not.

Comment from Jean-Marc Liotier on 11 March 2014 at 10:55

Isn’t a “report button” the obvious solution ? It requires moderators to act on the reports, but that is till faster and overall less work than the alternatives.

Comment from TheFive on 11 March 2014 at 22:32

If the majority is blocked in a short timeframe: Is it possible to delay publishing the blogs in Twitter about 2-4 hours ? (Or any other reasonable time) The OSMBlog Timeline in Twitter does not look really pretty.

Christoph

Log in to leave a comment