Commons:Bots/Requests/BotAdventures

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Operator: Ghouston (talk · contributions · Statistics · Recent activity · block log · User rights log · uploads · Global account information)

Bot's tasks for which permission is being sought: Adds files to a category in Category:Photographs by camera manufacturer based on Exif data. The intention is not to process millions of files, but to populate categories that don't have many entries.

Automatic or manually assisted: automatic

Edit type (e.g. Continuous, daily, one time run): Intermittent. Will probably select files from particular users or categories.

Maximum edit rate (e.g. edits per minute): 30-60, with Maxlag=5

Bot flag requested: (Y/N): Y

Programming language(s): Go, using go-mwclient.

--ghouston (talk) 11:59, 15 January 2015 (UTC)[reply]

Discussion

The proposal at Commons:Wikidata? This seems like a great idea, but it doesn't yet cover replacing categories. That would require a new user interface to allow selecting images by property and intersecting matches with the category system, and it may take a long time before it happens. --ghouston (talk) 23:38, 15 January 2015 (UTC)[reply]
EugeneZelenko, do you know if there is a public discussion going on anywhere with ideas for Wikidata-like functionality Commons? If there is I would be interested to read it. --MichaelMaggs (talk) 08:19, 25 January 2015 (UTC)[reply]
I linked the wrong page somehow. It looks like Commons:Structured_data is the current version of the project. The timescale is "next months and years", but if taken to its logical conclusion a large number of categories could be replaced by properties (anything with a date or file format for example). --ghouston (talk) 22:46, 25 January 2015 (UTC)[reply]
Well, the logical conclusion would be to replace all categories by properties, but presumably it will start with the easier cases. --ghouston (talk) 04:37, 26 January 2015 (UTC)[reply]
  • Where it may be useful is in the development of a translation table. It would be possible to copy the two Exif fields (camera manufacturer and model) to properties, but probably more useful to translate them to a single property that represents the camera model uniquely. There may be cases where the model values aren't unique among manufacturers, such as File:Big_crane_(3402145416).jpg has just "CX1". There are also cases where the model numbers are just ugly, such as "<Digimax S500 / Kenox S500 / Digimax Cyber 530>" in File:Abejorro_02.jpg. An idea would be to store a translation table in a wiki page somewhere, which would map the manufacturer/model combination to a single string, which would be "Ricoh CX1" and "Samsung S500" in these cases. Initially this bot would use the table, but it would later be available for a bot that sets properties. --ghouston (talk) 07:18, 14 February 2015 (UTC)[reply]
I've only run it manually on selected user uploads or categories. It could be done more systematically, perhaps using API:Allimages, which would be a good way to find recent camera models. Editing millions of files can be avoided by not adding to categories that already have more than say 500 members. --ghouston (talk) 00:57, 6 June 2015 (UTC)[reply]
Perhaps 100 in a category would be enough. The biggest proliferation of camera models now occurs in mobile phones, where manufacturers seem to spam new models constantly. --ghouston (talk) 01:03, 6 June 2015 (UTC)[reply]
Just for understanding: In which way is a category helpful that consists of semi-randomly selected images, a small percentage out of those that would actually fit into the category? --Krd 18:49, 6 June 2015 (UTC)[reply]
It gives a sample of images taken in real-world conditions from the particular camera. It would be more useful if all images were categorised, but the potential Wikidata / structured-data change seems to have made this undesirable. --ghouston (talk) 00:38, 7 June 2015 (UTC)[reply]
Sounds reasonable to me.  Support --Krd 06:56, 7 June 2015 (UTC)[reply]
It sounds reasonable to me, too ( Support) but I still don't see the advantage of not categorising all images.--Pere prlpz (talk) 22:42, 7 June 2015 (UTC)[reply]

As there are no objections, I'm going to approve this request. Bot flag granted. --Krd 18:47, 14 June 2015 (UTC)[reply]