Automated edits/confusedbuffalo/DfE establishments

Draft proposal, not yet proposed, website not yet ready, documentation not complete

In short, the proposal is to take open data from the Department for Education (DfE) relating to schools and colleges and use it to update data on objects in OSM. This will then be a guided import, with proposed edits being presented on a website and all edits being reviewed by a human.

Who?

I, confusedbuffalo am organising the data for the import and making the proposed edits available at https://dfe-to-osm.pages.dev/

Contact me by OSM message or comment on the forum post if you have any questions.

Once discussion on the forum has concluded, anyone is free to make edits using the website.

Motivation

Having up to date information on all objects in OSM is important. Schools are one such category which the UK community has a long history of caring about.

We can ensure that school names, contact details and more are up to date by comparing OSM data to official DfE data.

Applying some tags in a semi-automatic manner makes it easier to keep data up to date and saves mappers' time.

Description

Every school (and college and university and local authority-run nursery) in the UK is required to be registered with DfE and provide certain data. Each establishment has a Unique Reference Number (URN) which is tagged in OSM as ref:edubase, since the Get Information About Schools (GIAS) system used to be called Edubase. An example would be a URN of 1000000, the data about which can be seen here and which is tagged on The Aldgate School, w116465152.

Data can be downloaded from the GIAS site in bulk here, this includes data for school groups, such as federations or academy trusts.

Object Selection

amenity=kindergarten or amenity=school or amenity=college

Objects are matched to the source data using the value of the ref:edubase tag.

Unmatched objects are displayed for information, that is those without ref:edubase or those where the value of ref:edubase does not match an open school.

Kindergartens without ref:edubase are not included in the unmatched list, nor are objects with school=shared_site or college=adult_education, aligning with Robert Whittaker's https://osm.mathmos.net/schools/

Multiple-site schools

Some comments below relate to multi-site schools. There exist some schools which are legally one school, but operate at multiple sites. These are detected either if the object is a site relation or if school:multi_site=yes is tagged.

An example is Leadgate Primary School (URN 114039) which has the Infants and Nursery at one location w997789805 and the rest of the school at a different location w453786468. In this example, both sites are quite close and the name on GIAS gives a hint about it being split site, but this is not always the case.

(Semi-) Automatic Edits

The proposal is that the following tags can be automatically added or updated from the source data. This would be done by having a page that shows the proposed changes for a given region and providing an OSM file for download that contains these edits that can be verified and uploaded to OSM.

• ref:edubase
  • Only changed if the establishment has a 'successor' establishment in DfE data AND the only other changes are in this list, and none in the second list, below. Otherwise, it is marked for manual review
• ref:edubase:group
• operator
• operator:type
• operator:wikidata
  • Using a lookup table from ref:edubase:group to the wikidata tag, if it exists
• capacity (only added/changed if the school does not operate multiple sites)
• religion
• denomination
• diocese
• school (only added/changed if the school does not operate multiple sites)

• school:type
• school:gender
  • Only added for male/female
• school:boarding
  • Only added for yes
• school:selective
  • Only added for yes
• school:group:type
• check_date
  • Only added/changed if other changes are being made
  • The source data has a date of when the information was last updated or confirmed
  • Only changed if this date is more recent than the existing tag

Manually Reviewed Edits

The proposal is that if the data suggests changes to any of the following tags that no edits be made automatically, but that the whole element be reviewed by a human and the suggested edits adjusted as appropriate.

• name
  • Changes are not suggested if there is only a trivial difference between the names, including 'and' to '&' or variants of 'CE', 'C of E', 'Church of England' etc. (e.g. if the only difference is e.g. "Some CE School" and "Some C of E School" then no change is suggested)
  • If the name in the official data is already tagged in alt_name, official_name, name:en or name:cy then no change is suggested
• amenity
  • Either school, college or kindergarten
  • The duplicated education tag is updated at the same time, if already set. If it is not set then it is added only when there are other modifications for an object.
• addr:postcode
  • If the target element has not:addr:postcode then that is checked and no change is suggested if the new postcode matches that value
  • Only suggested if the school does not operate multiple sites
• ref:GB:uprn (Unique Property Reference Number)
  • Only suggested if the school does not operate multiple sites
• min_age
  • Only suggested if the school does not operate multiple sites
  • If the age is only changing by 1 then this would be in the automatic category
• max_age
  • Only suggested if the school does not operate multiple sites
  • If the age is only changing by 1 then this would be in the automatic category
• isced:level
  • A change to this would be made automatically if the min_age and max_age in the data match what is already tagged, due to a recent discussion on the forums which revealed that previous guidance on this on the wiki was incorrect.
  • This is calculated using min_age and max_age
  • Only suggested if the school does not operate multiple sites
• phone
  • Or contact:phone, respecting what is already tagged on the element
  • Phone numbers from GIAS are verified using the phonenumbers library and invalid numbers will not be suggested
  • Formatting changes (addition or removal of spaces) are not suggested
  • If the phone number is already mapped in operatorːphone then no change will be suggested
• website
  • Or contact:website, respecting what is already tagged on the element
  • Websites from GIAS are checked and redirects followed to obtain the final URL.
  • If the website on GIAS does not resolve to a valid site then no change is suggested
  • If the only change is to upgrade the OSM tag from http to https this would be in the automatic category
  • If the phone number is already mapped in operatorːwebsite then no change will be suggested

Removed tags

The following tags would be removed from objects where other changes are already being made, in either manual or automatic suggestions.

• school:trust
  • A yes/no value which is succeeded by information in school:group:type
• school:trust:type
  • As above, the information is better recorded in school:group:type
• school:trust:name
  • A previously criticised tag where the information is better contained in operator
• school:federation:name
  • Same as above
• school:age_range, age_range, source:school:age_range
  • Information is contained in min_age and max_age
• url
  • website or contactːwebsite is used instead
• postal_code
  • addrːpostcode is used for POIs

Consultation

Forum post tbc

Licence

Data from Get Information About Schools is available under the Open Government Licence v3.0 (wiki page). DfE have confirmed that this includes addresses and UPRNs in the data.

When?

Ongoing availability of the website from where edits can be made, after discussion has concluded on the forum.

Changeset Tags

The following changeset tags will be included in files available for download and in any edits made on the website itself.

source = https://get-information-schools.service.gov.uk/; https://dfe-to-osm.pages.dev/

discussion_before_edits = TBC

osm_wiki_documentation_page = https://wiki.openstreetmap.org/wiki/Automated_edits/confusedbuffalo/DfE_establishments

data_licence = Open Government Licence v3.0

For the automatic OSM edit files, the following tags would also be added

mechanical = yes

comment = Automatically update data on UK education establishments using Get Information About Schools Open Data

For the edits made on the website, the default changeset comment, which is editable, would be

comment = Manually check and correct suggested tags from Get Information About Schools Open Data

Opt out

If changes are being suggested by this tool that are clearly wrong and are an error in the tool then contact me and I will endeavour to fix it.

If changes are being suggested that are clearly wrong and are an error in the source data then you can try contacting the school in question and politely requesting that they update their record. Schools are legally obliged to keep their record up to date.

If the name is clearly not what should be tagged in OSM, then adding the name from GIAS in official_name will stop the change from being suggested.

If the postcode is incorrect (which can happen) adding not:addr:postcode will stop the change from being suggested.

If the school operates multiple locations then adding school:multi_site=yes will stop changes being suggested to addr:postcode, ref:GB:uprn, capacity, min_age, max_age, isced:level and school. The advice above about name still applies in such situations, or if the name from GIAS is part of the OSM name then no change will be suggested to it (e.g. no change will be suggested for "Some Primary School - Junior Campus" if the name in GIAS is "Some Primary School")

If there is an element that you do not want to display on the tool for some reason then discuss with me and I would be able to exclude it.