Re: [Talk-us] SEO Damage to OSM

[Clifford Snow]

I'm looking for input on doing a post on the Reddit/r/SEO subreddit
describing some of the damage done by a few and why adding their clients to
OSM won't help. What concerns me is that it will create
another TheSilphRoad where every Pokemon player wants to add fake water
features. Do you think it would be better to get the issue out in the open
or hope that obscurity prevails?

Clifford

-- 
@osm_seattle
osm_seattle.snowandsnow.us
OpenStreetMap: Maps with a human touch
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[John F. Eldredge]

The "mechanical Turk" term is not an ethnic slur, but instead an allusion 
to a famous 18th-century chess-playing automaton, made to resemble the 
upper body of a man in traditional Turkish clothing, mounted on a cabinet. 
It was eventually revealed to be a fake automaton, operated by a man hidden 
inside. The modern term refers to a human working a repetitive job, because 
they are cheaper than developing a computer program to do the job.






On June 30, 2017 6:23:15 PM Frederik Ramm  wrote:


Hi,

On 06/30/2017 07:24 PM, Ian Dees wrote:

Every edit was made by a new/separate user account that only ever makes
one edit. In most cases, these edits are useful (it's someone adding a
business POI) but in some cases they add the details to the wrong piece
of geometry.


I see. We had a couple similar issues in the past, presumably from some
kind of SaaS SEO tool that will automatically sign up new accounts for
users of the tool. But we might also be dealing with "mechanical
turk"(*) type of human work.

These tools tend to get more sophisticated in flying under our
collective radar, but usually not sophisticated enough to get the
tagging quite right and avoid adding duplicate data.

The addition of advertising copy in the changeset comment is something
I've seen a lot (often duplicated or amended by a note or description tag).

I wonder if downloading a changeset planet and feeding all changeset
comments to some sort of Bayes filter could help identify more problems.

Bye
Frederik

(*) Where I live this term would be considered really offensive towards
those who do this kind of work but it seems to be the normal term in the US?

--
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us




___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Erie Canalway Trail

[Kevin Kenny]

On Wed, Jul 5, 2017 at 12:54 PM, Kevin Kenny
 wrote:
>  In my neighbourhood in Niskayuna, the trail as placarded
> leaves the Mohawk-Hudson path at Blatnick Park, backtracks along River
> Road to Riverdale Road (there's a bike path on the shoulder and a
> signed crossing), and then continues on Chestnut Lane, Briar Ridge,
> Windsor Drive, Nott Street Extension, Orchard Park Drive, Morrow
> Avenue, and Niskayuna Drive. I only recently noticed the signage over
> in Old Niskayuna, and haven't yet gone over there to figure out how it
> gets back to the towpath. The path shown on the web site, which stays
> on the bike path all the way into downtown Schenectady, makes more
> sense to me, but it is NOT what is placarded! The part of the bike
> path that I use on my daily walk to work is not placarded for the
> Canalway Trail.

Egg on my face.

I misinterpreted the placards! They are marking a whole series of
approach routes TO the trail, rather than the trail itself, which in turn
explains why I found it so hard to make them into a coherent whole.

What was especially confusing is that, following the approach
routes, which are signed in only one direction, there's actually
no sign or other indication that you've got to the trail. You just come
out on River Road, with the (unsigned) Mohawk-Hudson
Bike-Hike Trail on the other side.

(I still do think we should have a route relation for the whole thing,
though.)

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Clifford Snow]

Can we curtail the discussion on which country is responsible? It doesn't
seem to help move us forward to tackling the problem.

I've met with two different SEO companies looking to add their clients to
OSM. It appears they didn't want to go through the import process nor did
they start adding companies. That's the good news. One described how easy
it was to upload to Google, Foresquare, Yahoo, Bing, and Apple. Just need a
csv spreadsheet using the proper business classification for the business.
(Each of the search companies have different requirement for
classification.) The result is when you search for these companies
locations, they are often geocoded to the street since there isn't an exact
location.

SEO want to add their clients to OSM because they think it will increase
Page Ranking. I was told that OSM has a very high Authority rank which
means that their client webiste benefits from being on OSM. What they don't
realize is when we display their URL  a ref=nofollow tag which tells the
search engines not to count that link when determining Page Rank. As was
pointed out to me, since we make our data available to others, they many
not be following our nofollow policy.

The easy answer to stop SEO companies from spamming is to remove the
website URL tag. Of course that's not going to fly.

I suggest we encourage users of OSM data to also use the nofollow tag. If
you are using OSM data, check to see how the url is presented.

Clifford


-- 
@osm_seattle
osm_seattle.snowandsnow.us
OpenStreetMap: Maps with a human touch
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Ethan Nelson]

Just to bring a little perspective from the DWG (but speaking in a personal 
capacity):


The issue of SEO groups adding information first came up at the beginning of 
2016. We were able to revert a lot of the data for a while in cooperation with 
vigilant users. Blocks are useless in these cases because you are playing 
whack-a-mole with new accounts. Additionally, I compiled a list of some of the 
users to check with sysadmins, and it turned out all the accounts had the email 
domain and IP address outside of the US. As a result, we added some 
countermeasures to try and stave off the issue. The main characteristics on 
these edits were each account being named after a business and adding only that 
business (including the 'Keywords' key in their tags). What made detecting 
these edits difficult at the time was mappers coming by and correcting the tags 
to be more "mainstream" (such as changing 'Keywords' to 'keywords' or 
'description').


Over time, though, the editors became smarter (attackers adapt to defenses) as 
they started to use "mainstream" tags instead of sore thumbs, register with 
emails that were not all the same, and register from different IP blocks. 
Without a clear connection, it was a matter of loose-leaf accusations against 
some company for accounts that blended in a little more. And I doubt the 
sysadmins would agree to block whole range of IP addresses.


Some of the edits shared a common thread and pointed towards an SEO company 
based in California. Over time, I have attempted to contact them. This has 
entailed sending emails to their support addresses (not only from my personal 
address but also from a DWG address), the CEO directly through their address on 
LinkedIn, filling out contact forms on their website, and tweeting them. In 
addition, I have tried to directly contact some of the businesses whose 
information was added to inquire who they hired for their online presence, 
website, and so on. Not a single response was received in any of these cases. 
And were I to try and call that company, I speculate I would not even be able 
to get a hold of anyone past level 1 support or sales.


Really, I can't say for certain this company is even responsible, so it's 
difficult to craft some official letter from e.g. the Board. It could be this 
company contracts out geoSEO work. And maybe even their contractor subcontracts 
it to yet another firm. Also, some of the examples show different 
characteristics than the first group I had brought up before (which I don't 
believe I ever saw add information to a street), so there could be more than 
one group at play here.


It is not the easiest task to automatically sort out legitimate businesses from 
SEO editors, so that means it will probably require manual screening of all new 
user accounts.

What's the solution? I don't know. That and vandalism detection are two million 
dollar questions (reward money not supplied).


Best,

Ethan aka FTA, speaking about his DWG experiences in a personal capacity

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Denis Carriere]

Great research Frederik!

*~~*
*Denis Carriere*

On Wed, Jul 5, 2017 at 5:05 PM, Frederik Ramm  wrote:

> Hi,
>
> > These spam changes do not need that complexity to detect.
>
> I've done some numbers, maybe it helps.
>
> I counted all users that only ever commited one changeset with one edit
> inside. This number is 140352.
>
> Then I discarded those where the changeset comment was shorter than 50
> characters or where the content had been redacted long time ago, leaving
> me with 12173.
>
> Then I looked at the objects modified/created, and discarded all where
> the object had neither website, nor description, nor note tag. This left
> me with 3323 objects.
>
> Then I looked at the list and found a broad range of edits. Some, while
> having an advertising slant, seem a legit addition of someone's own
> business:
>
> user=Martin Merkur
> changeset=38362589
> comment=Our doors are always open.  Come and visit, taste our coffee,
> see what we do
> object=node 4103514010
> addr:city=Berlin;addr:housenumber=38;addr:postcode=
> 12435;addr:street=Elsenstraße;amenity=cafe;cuisine=coffee_
> shop;internet_access=no;name=passenger
> coffee;note=https://www.facebook.com/PassengerEspresso/;opening_
> hours=7:30-15:00
> Uhr;smoking=outside;website=passenger-coffee.de
>
> or
>
> user=otheryan
> changeset=13150739
> comment=Added in West Town Bikes as it is at the same address and has
> enough of its own activity that it needs to be recognized on the map.
> object=node 1585399965
> addr:housenumber=2459;addr:postcode=60622;addr:street=W
> Division;name=Ciclo Urbano/West Town
> Bikes;shop=bicycle;website=http://ciclourbanochicago.com/
>
> some look more SEO-y
>
> user=northcarolinahealth
> changeset=43324244
> comment=Updated Osborne Insurance Services at Raleigh, NC
> object=node 4474950186
> addr:city=Raleigh;addr:housenumber=5316;addr:postcode=27609;addr:state=NC;
> addr:street=Six
> Forks Road;hours=Mon-Fri
> :8.00AM-6.00PM;name=Osborne Insurance
> Services;phone=919-845-9955;suite=110;website=http://
> northcarolinahealth.org
>
> or
>
> user=blakemanhart
> changeset=43027180
> comment=Updated State Farm - Blake Manhart at Springfield, VA
> object=node 4456153164
> addr:city=Springfield;addr:housenumber=8322;addr:
> postcode=22152;addr:state=VA;addr:street=Traford
> Ln #B;name=State Farm -
> Blake Manhart;Owner=Blake
> Manhart;phone=703-992-9664;website=http://blakemanhart.com
>
> I had a look at trying to automatically match website and user name; 457
> of them actually contain the user name in the web site. but that is a
> too coarse check. I fear that it might be necessary to look through the
> rest manually to detect the dodgy ones.
>
> Of the 3323, 208 have a highway tag. But here it bites me that I took
> everything that had either note or description or website, because some
> of the edits with highway=* are legit and have a description/note where
> the newbie mapper explained what they did. 170 of the 208 do have a
> website tag, and finally, they *all* seem dodgy. (Interestingly it was
> not all ways - some highway=traffic_signals too!)
>
> I've run a revert on these 170 but the majority had already been fixed
> by others!
>
> That leaves us with a good 3115 objects to investigate. Many do clearly
> violate our "no advertising" rules but then again we don't want to bee
> to harsh with the cycle shop owner who maybe oversteps the line.
>
> I've put my interim results here
>
> http://www.remote.org/frederik/tmp/username-in-url.csv
>
> (for those where the username is in the URL) - do you think we should
> revert them all automatically? (Keep in mind many may have been reverted
> already - we'd only work on those where the spam version is still current.)
>
> and
>
> http://www.remote.org/frederik/tmp/other.csv
>
> for those where the username is not (fully) in the URL.
>
> Bye
> Frederik
>
> --
> Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"
>
> ___
> Talk-us mailing list
> Talk-us@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us
>
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Frederik Ramm]

Hi,

> These spam changes do not need that complexity to detect.

I've done some numbers, maybe it helps.

I counted all users that only ever commited one changeset with one edit
inside. This number is 140352.

Then I discarded those where the changeset comment was shorter than 50
characters or where the content had been redacted long time ago, leaving
me with 12173.

Then I looked at the objects modified/created, and discarded all where
the object had neither website, nor description, nor note tag. This left
me with 3323 objects.

Then I looked at the list and found a broad range of edits. Some, while
having an advertising slant, seem a legit addition of someone's own
business:

user=Martin Merkur
changeset=38362589
comment=Our doors are always open.  Come and visit, taste our coffee,
see what we do
object=node 4103514010
addr:city=Berlin;addr:housenumber=38;addr:postcode=12435;addr:street=Elsenstraße;amenity=cafe;cuisine=coffee_shop;internet_access=no;name=passenger
coffee;note=https://www.facebook.com/PassengerEspresso/;opening_hours=7:30-15:00
Uhr;smoking=outside;website=passenger-coffee.de

or

user=otheryan
changeset=13150739
comment=Added in West Town Bikes as it is at the same address and has
enough of its own activity that it needs to be recognized on the map.
object=node 1585399965
addr:housenumber=2459;addr:postcode=60622;addr:street=W
Division;name=Ciclo Urbano/West Town
Bikes;shop=bicycle;website=http://ciclourbanochicago.com/

some look more SEO-y

user=northcarolinahealth
changeset=43324244
comment=Updated Osborne Insurance Services at Raleigh, NC
object=node 4474950186
addr:city=Raleigh;addr:housenumber=5316;addr:postcode=27609;addr:state=NC;addr:street=Six
Forks Road;hours=Mon-Fri
:8.00AM-6.00PM;name=Osborne Insurance
Services;phone=919-845-9955;suite=110;website=http://northcarolinahealth.org

or

user=blakemanhart
changeset=43027180
comment=Updated State Farm - Blake Manhart at Springfield, VA
object=node 4456153164
addr:city=Springfield;addr:housenumber=8322;addr:postcode=22152;addr:state=VA;addr:street=Traford
Ln #B;name=State Farm -
Blake Manhart;Owner=Blake
Manhart;phone=703-992-9664;website=http://blakemanhart.com

I had a look at trying to automatically match website and user name; 457
of them actually contain the user name in the web site. but that is a
too coarse check. I fear that it might be necessary to look through the
rest manually to detect the dodgy ones.

Of the 3323, 208 have a highway tag. But here it bites me that I took
everything that had either note or description or website, because some
of the edits with highway=* are legit and have a description/note where
the newbie mapper explained what they did. 170 of the 208 do have a
website tag, and finally, they *all* seem dodgy. (Interestingly it was
not all ways - some highway=traffic_signals too!)

I've run a revert on these 170 but the majority had already been fixed
by others!

That leaves us with a good 3115 objects to investigate. Many do clearly
violate our "no advertising" rules but then again we don't want to bee
to harsh with the cycle shop owner who maybe oversteps the line.

I've put my interim results here

http://www.remote.org/frederik/tmp/username-in-url.csv

(for those where the username is in the URL) - do you think we should
revert them all automatically? (Keep in mind many may have been reverted
already - we'd only work on those where the spam version is still current.)

and

http://www.remote.org/frederik/tmp/other.csv

for those where the username is not (fully) in the URL.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Simon Poole]

Just to put the whole thing into perspective: the current run rate of
SEO fake accounts in the US seems to be reasonably low. I counted 5 for
the last 7 days, that is roughly 2% of new mappers in the US during
those 7 days.




signature.asc
Description: OpenPGP digital signature
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Simon Poole]

Am 05.07.2017 um 17:27 schrieb Greg Morgan:
>
>
> Am 01.07.2017 um 08:33 schrieb Simon Poole:
>
> ...given that we have a known US based SEO company that has created
> (literally) 1000s of such accounts (but with slightly less spammy
> edits), and "we" haven't taken any action, why should we in this case?

I suspect you are barking up the wrong tree, "we" includes everybody in
OSM, nobody seemed to be in the slightest upset about that when I
published the diary post and nobody seemed to be inclined to run after
the SEO  company in question.

>
> How do you know this is all coming from a US based Simon?  How does
> targeting US locations make it an US based company?  If it the same
> occurred in Europe, then I am sure you would have more care and
> concern than you are showing here.  Dude I often cringe when I reas
> what you write in email.
>

Because I manually checked a largish number of the companies (the
subjects of the ads) last year and they had a trivial tell tale
attribute that pointed (as in 100% sure) to an US based SEO company.
That doesn't imply that the accounts that triggered this thread (which
were created before we saw the big influx in 2016) are from an US
company and I didn't say anything of the sort. Just that we didn't go
after an identified non-offshore offender at the time, so there is no
point in suddenly getting upset now about others playing the same or
similar games.

The other point is that we haven't seen a similar increase in fake
accounts in Europe, not that we don't have SEO companies here, but they
seem to operate differently, and indeed a number of them have made
contact with the local OSM communities first before adding anything (and
some haven't). 


signature.asc
Description: OpenPGP digital signature
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Frederik Ramm]

Hi,

On 07/05/2017 05:40 PM, Andy Townsend wrote:
> On 05/07/2017 16:27, Greg Morgan wrote:
>> I've seen the DWG go after real newbies because they are exited and
>> want to make a difference but make a few mistakes.
> 
> Have you got an example of that (offlist if it would be preferable)?

Yes, I would like to see that claim either substantiated or withdrawn
with an apology.

> A significant amount of my DWG time is spent trying to persuade mappers
> around the world to allow new users to make mistakes, which they
> inevitably will before they get the hang of things. I've certainly not
> seen "the DWG go after real newbies".

It does occasionally happen that the first thing an ambitious new
sign-up does is import a couple thousand nodes which will then swiftly
be reverted - but that's not "going after" someone, and I seriously
doubt that the freedom to sign up and import data without consultation
is what the community wants us to uphold.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frede...@remote.org  ##  N49°00'09" E008°23'33"

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Erie Canalway Trail

[OSM Volunteer stevea]

Hi Kevin:It may be that more than one relation is warranted because of multiple trail uses.  I can see how there might be one tagged route=hiking and another tagged route=bicycle, although if the trail is largely unpaved, route=mtb is preferred instead of the latter.  Of course, these would likely have to stay synced.  I don't know of another or better way to do this, as the route values do collide like this.  (The same is true for equestrians, if they are allowed on the trail:  "and then there were three.")Then there are the network= tag values.  Perhaps network=rcn is correct for the route=bicycle and network=rwn for the route=hiking.  For the route=bicycle, a cycle_network=* tag is a good idea.It may make sense to fit this into WikiProject_United_States_Long_Distance_Trails, too.Cheers,SteveACaliforniaFrom: Kevin Kenny Subject: [Talk-us] Erie Canalway TrailDate: July 4, 2017 at 10:07:57 PM PDTTo: "talk-us@openstreetmap.org" The Erie Canalway Trail is a historic corridor and a major cyclingroute that runs near the route of the Erie Canal from Buffalo toAlbany.  https://www.ptny.org/cycle-the-erie-canal is the web sitedescribing it, and there's an interactive trail map athttp://www.ptny.org/bike-canal/map/.Alas, the map is not always correct, at least according to the currentsignage. I know this because the trail runs through my neighbourhood.I've noticed the signs but not yet attempted to add the route to themap.Is anyone interested in getting a route relation started with thecorrect route?  I see in a number of other places in New York that thetrail is indicated by the name of a way, but the ways are not linkedinto a coherent route.I'm not a cyclist, but I'll happily sign up to field-check the signagenear me - I live only a city block or so from the marked trail.(Caveat: I may need someone else to make sense of what the trail mapis trying to tell me for the stretch from Cohoes to Watervliet. Itseems to be a bit of a maze there.)It could be that the route through my immediate neighbourhood is adetour to avoid a couple of very steep hills on the route on the website. I can check in the next few days where the routes rejoin, andwhether both are signed.
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Erie Canalway Trail

[Kevin Kenny]

On Jul 5, 2017 10:25 AM, "Richard Welty"  wrote:

> you may want to consider a super relation. some parts of the Canalway trail
> are themselves named trails, for example the Mohawk-Hudson Bike-Hike Trail
> from Schenectady to Albany. i think i have set up a relation for this
> one already,
> i'll go back and check.
>

You did indeed set up a relation for the Mohawk-Hudson
Bike-Hike Trail.
https://www.openstreetmap.org/relation/306742

(In mapping around here, I've been trying to take a lot of my
cues from you.)

I think having an ordinary route, not a super-relation, would be
better.  In my neighbourhood in Niskayuna, the trail as placarded
leaves the Mohawk-Hudson path at Blatnick Park, backtracks along River
Road to Riverdale Road (there's a bike path on the shoulder and a
signed crossing), and then continues on Chestnut Lane, Briar Ridge,
Windsor Drive, Nott Street Extension, Orchard Park Drive, Morrow
Avenue, and Niskayuna Drive. I only recently noticed the signage over
in Old Niskayuna, and haven't yet gone over there to figure out how it
gets back to the towpath. The path shown on the web site, which stays
on the bike path all the way into downtown Schenectady, makes more
sense to me, but it is NOT what is placarded! The part of the bike
path that I use on my daily walk to work is not placarded for the
Canalway Trail.

There's nothing wrong with having a way that's a part of multiple
route relations, and that's the only way to handle cases like this (or
like a great many of the long hiking trails; the New York Long Path
uses part but not all of such trails as the Peekamoose-Table Trail,
Burroughs Range Trail, Devil's Path, and Escarpment Trail in the
Catskills; the Appalachian Trail uses a couple of disjoint sections of
the Ramapo-Dunderberg Trail in Harriman Park, and so on.)

A super-relation may be needed from the point of view of data
management, because the number of member ways gets awkward to edit
otherwise. That's what I did with the Long Path
https://www.openstreetmap.org/relation/919642 - in that case I
somewhat arbitrarily decided to break up the route at county
lines. The handful of edits you'll see using Meerkartor from my user
name came about during the process of splitting it, because JOSM kept
crashing trying to deal with the huge relation. The bug wasn't in
JOSM, but in some sort of Java accessibility interface, and it's since
been fixed. In any case, I decided that if any of the tools struggled
that much, it was time to break things up.

I did not do the same thing with the Northville-Placid Trail
https://www.openstreetmap.org/relation/4286650 because the remoteness
of the country meant that I had fewer, longer ways to deal with and
didn't have to try to manage an unruly relation.
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Rihards]

On 2017.07.05. 18:40, Andy Townsend wrote:
> On 05/07/2017 16:27, Greg Morgan wrote:
>> I've seen the DWG go after real newbies because they are exited and
>> want to make a difference but make a few mistakes.
> 
> Have you got an example of that (offlist if it would be preferable)?
> 
> A significant amount of my DWG time is spent trying to persuade mappers
> around the world to allow new users to make mistakes, which they
> inevitably will before they get the hang of things. I've certainly not
> seen "the DWG go after real newbies".

i would like to express great gratitude and admiration at how dwg has
been handling things for the last few years.
you guys are awesome in patience, attention to detail and overall
quality. thank you.

> Best Regards,
> 
> Andy-- 
 Rihards

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Andy Townsend]

On 05/07/2017 16:27, Greg Morgan wrote:
I've seen the DWG go after real newbies because they are exited and 
want to make a difference but make a few mistakes.


Have you got an example of that (offlist if it would be preferable)?

A significant amount of my DWG time is spent trying to persuade mappers 
around the world to allow new users to make mistakes, which they 
inevitably will before they get the hang of things. I've certainly not 
seen "the DWG go after real newbies".


Best Regards,

Andy



___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] SEO Damage to OSM

[Greg Morgan]

On Fri, Jun 30, 2017 at 10:23 AM, Frederik Ramm  wrote:

> Hi,
>
> On 06/30/2017 06:21 PM, Clifford Snow wrote:
> > Edits, from what appears to be a search engine optimization company
> > (SEO), have damaged a number of ways in the US.
>
> Was it not possible to determine the user accounts responsible and then
> have DWG revert all their contributions? Or did these accounts also
> contribute good data?
>


Back in April I did send an email to DWG. I also included the response with
the DWG name redacted[1].  Here's part of the problem.  I have identified
an account that is a spam account.  There is a pattern to these spam
edits.  If I do tell you that only bad things come from a single account
and I am not engaging in an edit war and if I tell you the way that was
altered was malicious, then to leave these spam accounts around, let's them
still win.  They have their account with the same data as in the tags.

Am 01.07.2017 um 08:33 schrieb Simon Poole:

...given that we have a known US based SEO company that has created
(literally) 1000s of such accounts (but with slightly less spammy edits),
and "we" haven't taken any action, why should we in this case?

How do you know this is all coming from a US based Simon?  How does
targeting US locations make it an US based company?  If it the same
occurred in Europe, then I am sure you would have more care and concern
than you are showing here.  Dude I often cringe when I reas what you write
in email.

The ChanlerAir is the more pernicious version of the problem. This pattern
changes a real OSM feature such as a road. OSM Change Set Analyser may be
useful here because they changed the street name.

In the second version of the problem[2], the change is subtle but uses some
of the same principles. LOL, I included the email that I sent to the Tax
firm.  You can email the OSM account and any of the firms involved but you
will not get a response.

Bayesian spam filters have been used for years to work on spam issues.
These spam changes do not need that complexity to detect.
 1.) The OSM account name has something to do with the change: Copper
Canyon Law LLC Tax Attorney
 2.) I cannot confirm this but I am guessing that the account name is the
same or similar to the email address used to setup the OSM account.
 3.) There is only one edit.
 4.) There may be a gravitar on the OSM account.  Think about it.  How many
newbies have a gravitar and make one edit?
 5.) Most of the time, The OSM account profile page, the change set text
and description tag all have the same business oriented marketing text.
The changeset text and the description tag look the same in the cases that
I have observed.
 6.) If there is a name tag, then it may be the same as the OSM user
account.  This is where the pernicious version of these changes can be
detected because the street name is changed.
 7.) The addr:.* tags are abbreviated.
 8.) The addr:state may not be the in the same location as where the edit
was made i.e. the location is AZ but addr:state is MD, WA, CA.  The rub
here is that most of the change set tools treat all of the US as one area
verses the need for multiple areas.  If I see a newbie make a change via
Pascal's tool,  I cannot tell if a new change in MD is correct or not
because I live thousand of miles away.
 9.) The actual change will have a full email address.  I have not seen
many newbies master that many tags on their first change.
10.) There will be a phone number.
11.) As stated already, there will be a description tag.  In my case, I use
description tags on relations but not in all cases.  The description tag
may be one of the key markers for this kind-of spam edit.
12.) There will be an email address.
13.) There may be a fax number.
14.) There will be a website.
15.) In the pattern of nodes that I present, I show that open areas and
strange locations are involved.  LOL we do a lot with our washes in Arizona
because we can put a golf course where the water runs infrequently.  We
will not put an Asian eatery nor a tax law firm by a flood control feature.



> Was it not possible to determine the user accounts responsible and then
> have DWG revert all their contributions? Or did these accounts also
> contribute good data?
>

This is where I think that the DWG needs an oversight board. I've seen
where the DWG will rollback good edits like the recent wikidata rollback
without community input.  I've had to cleanup these crappy reversions. I've
seen the DWG go after real newbies because they are exited and want to make
a difference but make a few mistakes. However, when there is a clear
pattern, the attitude is that the DWG all will not lift a hand.  That's the
perception that I have of the DWG.

By the way, Fred, I am not asking you to delete these nodes.  I have been
saving the examples for other mappers and perhaps a tool like OSM CA can
make a go of detecting these malicious edits.

Regards,
Greg

Re: [Talk-us] FBI using OSM on website... without attribution

[Mike Thompson]

On Thu, Jun 29, 2017 at 2:14 PM, Mike Thompson  wrote:

> https://www.fbi.gov/contact-us/field-offices/denver
>
> See map in upper right part of page
>
Looks like they fixed it!   Thanks for everyone's help.

Mike
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Erie Canalway Trail

[Richard Welty]

On 7/5/17 1:07 AM, Kevin Kenny wrote:
> Is anyone interested in getting a route relation started with the
> correct route?  I see in a number of other places in New York that the
> trail is indicated by the name of a way, but the ways are not linked
> into a coherent route.
you may want to consider a super relation. some parts of the Canalway trail
are themselves named trails, for example the Mohawk-Hudson Bike-Hike Trail
from Schenectady to Albany. i think i have set up a relation for this
one already,
i'll go back and check.

richard

-- 
rwe...@averillpark.net
 Averill Park Networking - GIS & IT Consulting
 OpenStreetMap - PostgreSQL - Linux
 Java - Web Applications - Search


___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us