[OSM-talk] Better Ωuality monitoring needed from organized projects : Was Flagging issue with Micron organised editing
pierzenh at yahoo.fr
Wed Apr 14 16:35:59 UTC 2021
Pete Masters started a previous thread about poor organized mapping by one company. See https://lists.openstreetmap.org/pipermail/talk/2021-April/086444.html
I start an other thread to discuss more largely about Organized editing quality problems and insufficient monitoring / correction.
This Micron participation to a HOT tasking Manager project is not an isolated problem. For https://tasks.hotosm.org/projects/10110 project, there are other contributors with quality problems and no Validators yet. Note that no. of tasks validated is not always a good indicator of quality. And there is some limit to wha validators can do. Organizers should assure quality. We also need synthetic quality monitoring tools that alert about problems for specific projects or group of contributors.
Since HOT and Missing Maps partners have intensified mapping projects from 2014, we constantly report problems and never progress as a global community to evaluate the extent of quality problems and find solutions to correct.
For the OSM Nepal 2015 earthquake Response, we had peaks of contributions with more then one million objects a day. Validators could not sustain and alarmed us of the poor quality. I think that we can relate this with the Missing Maps project that started end of 2014 and the intensification of Mapathons. This Presentation (https://www.slideshare.net/pierzen/kll-workshop-2015-0809-openstreetmap-response-2015-nepal-earthquake ) shows contributor profiles For the Nepal Response that probably reflect participation in general.
At the end of 2016, we did have the same problem with the OSM Matthews response for south-west Haiti. When Severin Menard alarmed about the bad quality mapping, you should read the answers of some HOT supporters saying bad quality is not such a problem in the context of Crisis Response
(https://lists.openstreetmap.org/pipermail/hot/2016-October/012536.html). Badly the Haitians contributors have survival problems and not the capacity to repair this. Bad quality is still there.
The DRC Response for the North-Kivu Ebola emergency at the end of 2018 had to rapidly resolve problems of bad quality with organized mapping. Alerted by the field team in Butembo, we had to erase and restart rapidly the mapping of the city this in a major Epidemy + Security crisis. See this discussion where I also talk about quality mapping tool that I developped. https://lists.openstreetmap.org/pipermail/hot/2018-December/014662.html
Since 2015, major partners like MissingMaps have extended vaccinations Mapping projects covering large territories, moving constantly to new areas. They have also systematically covered areas of Disaster responses with Tasking manager projects this with often no coordination with the volunteer teams organizing responses.
With the intention from partners to ever intensify such projects, it might be time to have a look backward, analyze the situation and correct, and find ways to collaborate with openness and respect.
Le mercredi 14 avril 2021 09 h 42 min 00 s UTC−4, Pete Masters via talk <talk at openstreetmap.org> a écrit :
Hi all, writing with my HOT Head of Community hat on...
Just want to flag that there has been a large scale mapping effort done by a company called Micron, which has led to some really poor quality mapping and damage to data mapped previously by communities in different places…
Micron have (with good intent but little knowledge) jumped into mapping as an organisational activity both through the HOT tasking manager and by directly editing OSM. The edits can be identified through Pascal Neis’ changeset tool (#gomicron21 changeset comment).
These edits were brought to HOT’s attention on Monday and we have taken the following steps:
Found contacts at Micron and started a dialogue (they have now asked all employees to stop mapping)
Notified the DWG of the situation and sought advice on how to help resolve the poor quality edits
Mobilised validators to support cleanup of map data (using #microncleanup to identify these changesets)
Reaching out to communities where we have identified the biggest impacts
I am posting here to flag that this has happened, but also to respond to a changeset comment and discussions happening on the talk-gb channel around another project using the tasking manager that has been affected.
The Micron mapping was not done in collaboration or with support from any other organisation or community (including HOT or other Missing Maps members) and the company did not share their plans prior to starting to edit. In fact, HOT had no awareness of the company before researching the changesets once they were brought to our attention.
However, members of HOT staff are working to help resolve the situation and appreciate the support we are receiving from the wider OSM community.
A big thanks to those that spotted the issue and raised it, and to all those helping with data validation and correction!
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk