User:Crouch, Swale/Bot tasks/Civil parishes (current)

This is for civil parishes in England. Example - Rattlesden.

Rattlesden is a settlement and civil parish in the Mid Suffolk district, in the county of Suffolk, England. The parish also includes the settlements of Hightown Green and Poystreet Green. The settlement is located 15 mi north west of Ipswich. In 2011 the parish had a population of 959. The parish touches Buxhall, Shelland, Gedding, Drinkstone, Brettenham, Felsham and Woolpit.

Features
There are 59 listed buildings in Rattlesden.

Title

 * First tries in unqualified title, thus Rattlesden
 * Then "Placename, County", thus Rattlesden, Suffolk (Eg Walton, Cumbria)
 * Then "Placename, District, thus Rattlesden, Mid Suffolk (Eg Hayton, Carlisle)
 * If there are multiple in a district it uses coordinates, thus Rattlesden, 52.19, 0.89, this would probably need to be manually moved to a better title. Welburn is a case where this is needed.
 * It could also use "Placename (parish)" but that's probably unnecessary since there are few situations where there should be separate articles for parishes when they have the same name as a settlement, see User:Crouch, Swale/Civil parishes/Splits and Scotforth/Scotforth (parish) for example where this is needed but there are so few that they can probably be created manually.


 * Note that the checks for other topics when disambiguation is needed are with civil parishes and OS settlements only (other than other articles with that name on WP) thus it puts it at the base name if it is the only parish but disambiguates further than the county if there are OS settlements (even if they don't have articles). For example if it was creating Parton and it found it to be a red link it would create a page at the base name but as there is a DAB there it would see that disambiguation is needed but it would also spot that there is a Parton in the Allerdale district so it would create the article at Parton, Copeland.

Other ideas
See Fulford, Staffordshire for the textual example.
 * Geography and demographics like Danbury, Nebraskacompare.
 * Name origin from KEPN.
 * Domesday name.
 * Other population data (and similar) from Vision of Britain. Possibly in a table like Clare, Suffolk.
 * Weather like at Salta, Cumbria.
 * Location data (such as parish, district, postcode district, dialing code and post town).
 * Elevation.
 * Images from Commons.
 * Parish council links.
 * GENUKI and similar websites.
 * Stub tags and project tags (on talk).
 * Use Historic England instead of British Listed Buildings in which we search under "Parish (Civil/Non-Civil)" but this fails when there are 2 or more parishes of the same name in a district (such as Flixton).

Problems

 * Nomis often merges date for parishes of less than 100 into a larger one (see Swinhope for example and discussion at User talk:Bob Henshaw). Thus Thorganby doesn't have data and Swinhope has data for both parishes. Possible solution is using City population instead for the larger parish but that also doesn't publish data for the smaller one so working out the different between the Nomis figure and City population figure could be used for the smaller ones unless multiple small parishes are merged. Note that the area is still provided by City Population. so that could still be used. Surely the 2001 census data (like Newbiggin, Kirkby Thore) must exist somewhere.
 * Now using City Population to avoid the merges problem, it will still result in the small parishes not having population data but they will still have area.
 * Dealing with parishes that are redirects, some are alternative names such as Aston-on-Trent/Aston upon Trent where there should only be 1 article but some are redirects that should actually be separate articles such as Melbury Sampford redirecting to Melbury House and Brewood and Coven redirecting to Brewood. A possible solution is that I add a template informing the bot to create the redirects that I add it to and not otherwise create those that are redirects.

Process

 * A rate of around 10 articles created a day by the bot seems sensible, this is obviously very slow for a bot but this is probably quick enough especially if I am expected to check/improve each one. A faster rate of say 20 (or even 50 a) day could be done but I don't think it needs to be that fast even if I don't need to check each one.
 * Bot requests/Frequently denied bots specifies that mass creation of stubs are generally denied but that "exceptions do exist, provided the database contains high-quality/reliable data, that individual entries are considered notable, and that the amount of stubs created can be reasonably reviewed by human editors" in this case this seems to have been satisfied in addition to the fact that I will point this out to the WikiProjects.
 * Bots are far more efficient in producing high quality articles than humans and are less likely to make mistakes providing they are programmed correctly.
 * Many articles on places start out as sub stubs or articles with trivia so its not like bot creations are poorer than human ones. Any better content can be added if/when.
 * There is no need for the bot to deal with adding hatnotes/DAB entries etc since I will be manually checking articles that are disambiguated (see ) even if I am not expected to check all. I will check things like if I need to merge the article into its village if it is just an alternative name such as the Aston-on-Trent/Aston upon Trent example, if I need to add a hanote or DAB entry or if the article should actually be under the unqualified name in which I will need to use WP:RMT, this might be if the title redirects to say a nearby place, the title is a redirect to an unrelated article but the sub topic is NN in which the parish article should be at the base name or that the article should be moved to the base name because its primary.
 * There's probably no need to create Welsh Communities this way since there are only a few missing, see User:Crouch, Swale/Communities however if we're doing the adding content to existing articles they should be included making sure to use the term community instead of civil parish.
 * If the difference between an existing article and one are trivial the bot would create and the coordinates are in the same place then the bot instead creates a redirect for example Clyst St George/Clyst St. George and Shiptonthorpe/Shipton Thorpe. This is to reduce the number of duplicates that I would need to merge.