HOT Statistics

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

HOT Statistics

Mike Thompson
I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples). I am interested in things such as:
* How many people contributed
* How many features mapped (e.g. # of buildings, km of roads, etc)
* How quickly the contributions were made.

Are these available anywhere, or does anyone have a method for pulling them out via the API or Overapass?

Thanks,

Mike

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

john whelan-2
You might like to think in terms of data quality.  Number of mappers is interesting I've seen 20 new maperthon mappers not finish a single tile between them and I've seen a team of three JOSM mappers map a project very quickly doing about 90% of the project the other 10% was done by other mappers.

One of the end users once said to me at a conference OSM is great stuff but the quality is very uneven.

What can impact the rate of mapping is having a validator sitting on the project.  Giving feedback seems to motivate the mappers to do a few more.

Cheerio John

On 29 September 2016 at 18:22, Mike Thompson <[hidden email]> wrote:
I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples). I am interested in things such as:
* How many people contributed
* How many features mapped (e.g. # of buildings, km of roads, etc)
* How quickly the contributions were made.

Are these available anywhere, or does anyone have a method for pulling them out via the API or Overapass?

Thanks,

Mike

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot



_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Imre Samu
In reply to this post by Mike Thompson
>  I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples).


one  example: 
Typhoon Haiyan (2013)
- OpenStreetMap Activities for Typhoon Haiyan (2013)
http://resultmaps.neis-one.org/osm-typhoon-haiyan-2013-contributors  
 ( "To all 1,679 voluntary contributors of the OpenStreetMap project for Typhoon Haiyan (2013) who made more than 4,799,290 Map changes!"   )

Other: 
West Africa Ebola Response 
http://neis-one.org/2014/11/ebola-response-map/    "In total, more than 2,000 contributors made more than 10 million changes to the map"
     "To all 4,664 voluntary contributors of the OpenStreetMap project for  Ebola Response (2014) who made more than 18,074,848 Map changes!"

"
The statistics on contributor activity take account of nodes, ways and relations that were created, modified or deleted.
The net effect of these contributions on the OpenStreetMap database is remarkable. In comparison with the OSMPlanet of early 2013, the net effect on the map as of August 15 for Guinea, Liberia and Sierra Leone is :
    • 62,442 km of highways added from 21,597 (+289%)
    • 15,124 km of waterways added from 16,000 (+94%)
    • 11,162 Places added
    • 505,372 buildings added
"

and maybe useful: 
Crowdsourced Validation and Updating of Dynamic Features in OpenStreetMap An analysis of Shelter Mapping after the 2015 Nepal Earthquake 
http://wrap.warwick.ac.uk/78701/

Imre



2016-09-30 0:22 GMT+02:00 Mike Thompson <[hidden email]>:
I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples). I am interested in things such as:
* How many people contributed
* How many features mapped (e.g. # of buildings, km of roads, etc)
* How quickly the contributions were made.

Are these available anywhere, or does anyone have a method for pulling them out via the API or Overapass?

Thanks,

Mike

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot



_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Laura O'Grady
In reply to this post by john whelan-2
I agree with John. 

The issue in VGI (volunteered geographic information) or crowdsourcing has shifted from measures like how many people, number of edits and completion times is to one of measuring quality.

An important metric towards measuring this is credibility (is the information believable?). Two dimensions include whether the information is trustworthy (or good) and expertise (is the information source knowledgable?).

There are studies that are measuring the quality of OSM data in general by comparing edits by volunteer contributors to maps that have been created using validated sources.

Some research of this nature has been published in this book, which is available for free download at the link below:

Capineri, C, Haklay, M, Huang, H, Antoniou, V,  Kettunen, J, Ostermann, F and Purves, R 2016 European Handbook of Crowdsourced Geographic  Information. London: Ubiquity Press.


-
Laura O'Grady
[hidden email]


On Sep 29, 2016, at 9:26 PM, john whelan <[hidden email]> wrote:

You might like to think in terms of data quality.  Number of mappers is interesting I've seen 20 new maperthon mappers not finish a single tile between them and I've seen a team of three JOSM mappers map a project very quickly doing about 90% of the project the other 10% was done by other mappers.

One of the end users once said to me at a conference OSM is great stuff but the quality is very uneven.

What can impact the rate of mapping is having a validator sitting on the project.  Giving feedback seems to motivate the mappers to do a few more.

Cheerio John

On 29 September 2016 at 18:22, Mike Thompson <[hidden email]> wrote:
I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples). I am interested in things such as:
* How many people contributed
* How many features mapped (e.g. # of buildings, km of roads, etc)
* How quickly the contributions were made.

Are these available anywhere, or does anyone have a method for pulling them out via the API or Overapass?

Thanks,

Mike

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

joost
Hi,

Of course good research goes way beyond basic statistics. But basic statistics have their value too.


* You can get some nice stats for a custom area with this tool: http://osm-analytics.org/#/

* Missing maps has some road length statistics by country: http://www.missingmaps.org/osmstats/
(it is easy to demonstrate the possibly negative impact on map evolution of hot activations and imports. Compare for example road network length for Haiti and Dominican Republic. Mind you, there are many alternative explanations possible)

* The easiest way to get the data from a certain activation/mapathon is by using Overpass:
- this gives you all the changes in a bounding box within a certain time: http://overpass-turbo.eu/s/iJM (you can just move the map or draw a bounding box)

(and you can extract the polygon in json used for a certain task from the tasking manager)

- if map data becomes really big, you can download without visualising:

- you could also generate a csv list with just the stuff you need; here's an example: http://overpass-turbo.eu/s/iJO
(it should be possible to get the username into that list, but I didn't find how. The people at help.openstreetmap.org are extremely helpful with that kind of thing)


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Erno Mäkinen
In reply to this post by Mike Thompson
Hi Mike,

You might like to see these three examples:
* http://ernoma.github.io/mapathon/SPR/
* http://ernoma.github.io/mapathon/SPR_Sept2016/
* http://ernoma.github.io/mapathon/april2016/

The first two examples contain contributions from almost only mapathon
participants. The last one though includes statistics of contributions
from people all over the world because there were a lot of people
contributing after the earthquake in Ecuador.

The statistics are based on the Overpass API data collected in pretty
much the way as Joost Schouppe suggested. When you use "out meta" mode
in the query, it includes also usernames.

The code for above examples though not at all polished is also
available:
https://github.com/ernoma/ernoma.github.io/tree/master/mapathon 
(SPR_Sept2016 folder has the latest javascript file). Also, when
collecting the data from a large area, I strongly suggest to install
Overpass API on your own server.

BR,
--
Erno Mäkinen

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Mike Thompson
In reply to this post by john whelan-2


On Thu, Sep 29, 2016 at 7:26 PM, john whelan <[hidden email]> wrote:
You might like to think in terms of data quality.  Number of mappers is interesting I've seen 20 new maperthon mappers not finish a single tile between them and I've seen a team of three JOSM mappers map a project very quickly doing about 90% of the project the other 10% was done by other mappers.

John, I completely agree. For this presentation I am just trying to get the message across that humanitarian mapping in OSM is of significant size. 

What can impact the rate of mapping is having a validator sitting on the project.  Giving feedback seems to motivate the mappers to do a few more.
very good advice. 

Mike 

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Mike Thompson
In reply to this post by Imre Samu


On Thu, Sep 29, 2016 at 7:58 PM, Imre Samu <[hidden email]> wrote:
>  I am doing a presentation and I would like to cite some stats about some HOT projects (I just need three or four examples).


one  example: 
Typhoon Haiyan (2013)
- OpenStreetMap Activities for Typhoon Haiyan (2013)
http://resultmaps.neis-one.org/osm-typhoon-haiyan-2013-contributors  
 ( "To all 1,679 voluntary contributors of the OpenStreetMap project for Typhoon Haiyan (2013) who made more than 4,799,290 Map changes!"   )
....
Imre, these are all very helpful! Thanks.

Mike 


_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Paul Norman
In reply to this post by Mike Thompson

On 2016-10-01 03:56 PM, Mike Thompson wrote:


On Thu, Sep 29, 2016 at 7:26 PM, john whelan <[hidden email]> wrote:
You might like to think in terms of data quality.  Number of mappers is interesting I've seen 20 new maperthon mappers not finish a single tile between them and I've seen a team of three JOSM mappers map a project very quickly doing about 90% of the project the other 10% was done by other mappers.

John, I completely agree. For this presentation I am just trying to get the message across that humanitarian mapping in OSM is of significant size. 

The ballpark figure that I use is that HOT (the organization) + missing maps activity is about 5-15% of total activity. Humanitarian mapping is a bit broader, but probably wouldn't change the numbers much.

Exact figures depend on what you define activity as, but 5-15% is valid for

- Active mappers who have at least 1 HOT changeset
- Total mappers who have at least 1 HOT changeset
- Number of changesets
- New signups who have a HOT changeset

Unless a big focus of your presentation is on activity and how to measure it, I'd just stick with 10% as being approximately right.

_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot
Reply | Threaded
Open this post in threaded view
|

Re: HOT Statistics

Mike Thompson


On Sat, Oct 1, 2016 at 8:18 AM, Paul Norman <[hidden email]> wrote:


Unless a big focus of your presentation is on activity and how to measure it, I'd just stick with 10% as being approximately right.

Thanks Paul.  That is very interesting and I may cite it in my talk. I should have been more clear, I am not trying to show that humanitarian mapping is significant relative to the whole of OSM, just that a significant amount of mapping gets accomplished in response to these events.

Mike
_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot



_______________________________________________
HOT mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/hot