Commons:British Library/Mechanical Curator collection/map tag campaign
Launched with an all-day live event at the British Library on October 31st, Wikimedia and the British Library and OpenStreetMap are trying to identify all the maps in the Mechanical Curator collection, a set of 1 million images extracted from digitised 19th century books.
This page explains why, and how volunteers with an hour or so to spare can help identifying these maps. Once they have been identified we can locate them on the globe, and then upload them en-masse to Wikimedia Commons. Up to 10,000 may still remain in the British Library's photostream to be identified.
![](https://upload.wikimedia.org/wikipedia/commons/thumb/0/09/Tagathon_invitation.png/180px-Tagathon_invitation.png)
Tagging can be done online from anywhere in the world.
With enough online help, it would be great if we could find and tag 10,000 more maps by Monday 3 November
Nutshell[edit]
- "Gotta Tag 'Em All"
- The objective is to systematically use these index pages to go through the Mechanical Curator collection, and
- add the Flickr tag map to every image in the collection that represents a map or a ground-plan, and
- replace every Untagged maps? templated link on the index pages with either a maps link or the words no maps
- This will allow a new round of geo-referencing, which is the key to unlocking these images to make them findable and reusable - once maps have been identified and georeferenced, we will be able to bring them to Commons and make use of them. But not until we know where they are...
- The British Library launched the campaign with a Digital maps Halloween tagathon (RSVP), on Friday October 31st -- see below for full details. As well as the tagging, there were short interesting talks about various digital mapping activities, visits to the BL's map reading rooms, and free Halloween entry to the Library's atmospheric current major exhibition, Terror and Wonder: The Gothic Imagination.
- The maps are from all over the world and the tagging can be done from anywhere in the world. You can help. See the map tag status progress page for a current list of index pages still containing books that may have untagged maps.
Background motivation[edit]
- Examples of maps that have already been found and georeferenced can be found in this set on Flickr. Georeferencing means the same set of maps can now be plotted on a map of the world, making finding much easier, or laid over a modern map for comparison.
- More maps need to be found before there can be any more georeferencing. However, the British Library has to pay a fixed set-up fee for every round of georeferencing it starts, so it makes sense to try to get as many images as possible into each round -- all of them, if at all possible. It would be good to know that the Mechanical Curator collection had been systematically gone through, and we had found every one.
- Information from the geo-referencing is released by the British Library free, CC0. For WM, the information from a round of georeferencing is directly usable by us here on Commons -- for example in this category are some maps that we already have here, that have links on their description pages to see them in the BL georectifier. WM's own georeferencer is now live on labs, http://warper.wmflabs.org/, and the raw georeferencing information can be directly used in it.
- The project is also a great way to explore the Mechanical Curator collection (and our index to it), to discover the sort of things that are in it for different parts of the world. The big focus will of course be maps; but if anyone gets inspired to start tagging anything else they find, or to upload particularly useful pictures here to Commons, or to add them to articles, so much the better...
- It will be particularly valuable to identify and remove the books that have no maps from consideration; tag the maps in books that have only a few maps; and so the tags that remain to be added become progressively concentrated in big atlases, with fewer and fewer books left to consider. Such books with lots of maps are the easiest to mop up at the end, to close out the whole process.
Tagging process [edit]
The basic idea is simply to use the "Synoptic Index" that's on the Wikimedia site to go through the collection on Flickr, and tag all the maps.
To explain in more detail, it's useful to go to the map tag status progress page, and then choose one of the links there, for example this is the top of the UK and Ireland page.
Each entry on this page links to the collection of images on Flickr that has been scanned for each book. To this has been added
- a maps link to everything for that book on Flickr which has so far been tagged as a map (if in fact anything has), and
- a pink Untagged maps? link, to everything on Flickr for that book which has *not* (yet) been tagged as a map.
This has had the effect of turning the Synoptic Index page into a sea of pink from top to bottom.
The process of removing the pink links will then be one of
- Picking a book (i.e. a line in the index), and clicking through to the Flickr page for everything on that book that has not yet been tagged as a map
- If the book does contain some maps, add a maps link, ...
- ... then go through the Flickr page for that book, opening any images that are maps or ground plans and adding the Flickr tag map, ...
- ... and then, when there are no more maps on the Flickr page to tag, remove the pink "Untagged maps?" link
- Alternatively, if the book doesn't contain *any* maps, replace the pink "Untagged maps?" link on the Commons page with the words "no maps",
(Note that you need to be logged with a Flickr account to be able to add tags, which can be found on grey buttons below the image).
In this way the sea of pink should rapidly become moth-eaten, and then disappear altogether -- and in the process, we should have tagged all of the maps there are to find in the collection.
Additional tags[edit]
The following additional Flickr tags have been particularly requested by the British Library, in addition to the map tag:
- split -- if a single image contains multiple maps
- conical -- if the map is not on a simple rectangular projection
- world -- for maps which display the entire world on one sheet
- rotatec -- for images which need to be rotated clockwise
- rotatecc -- for images which need to be rotated anticlockwise
- rotate180 -- for images which are upside down
Wikicode[edit]
For somebody who knows wiki-editing, the process of making the changes required to the wiki page should be quite straightforward.
(Anyone actually in the room at the BL event won't in fact need to have to edit wiki at all, as discussed below).
Looking at e.g. the edit screen for the first section of the synoptic index page we were looking at,
- the pink Untagged maps? link is generated by the markup code
{{UntaggedMaps|sysnum=004159563}}
- the analogous bold maps link is produced by changing this (or copy-and-changing this) to
{{HasMaps|sysnum=004159563}}
while if there are no maps, simply remove the UntaggedMaps template altogether, and replace it with
no maps
For further questions, see immediately below.
No need to know wiki[edit]
Anyone can help without having to edit wiki at all as you can find maps and plans on FLickr and add map tags to each map image on Flickr, just leave one in each batch so that someone else can change the wiki tag for the batch to "HasMaps"
Questions[edit]
- What should be tagged ?
- All maps and plans that could be placed on a globe should be tagged. Some examples, including use of the additional
world
,conical
, andsplit
tags are shown here, but please include even quite simple outline maps, hand-drawn sketches, building plans, and 'medieval-style' maps with drawn buildings.
- All maps and plans that could be placed on a globe should be tagged. Some examples, including use of the additional
- Why won't Flickr let me tag ?
- To be able to tag, you first need to be signed in to Flickr with a Flickr account.
- Then if you follow an Untagged maps? link to the page of images on Flickr for the book, and then click on an image to open the specific page for that image, you should be able to scroll down the page and see a set of tags on grey button backgrounds, in the middle right of the screen, below the image. Just to the top of this set, on the right, should be the link "Add tags", and clicking on it should open a box to add new tags.
- Flickr says "Oops! There are no matches"
- This can sometimes appear, even when there should be matches. In such a case waiting a few seconds and reloading the page should bring up the search results expected. (Possibly Flickr was trying to serve a page from a just-in-time cache before the cache has been filled.)
- The message can also appear if there are indeed no returns for a particular search -- ie if a book only contains maps, and all of them have been tagged. (In which case the "Untagged maps" template should be removed). This can be checked by clicking on the book title link to show all its images, and seeing whether they are indeed all maps.
- If neither of these are the case, please leave a message on the talk page.
- What about non-map tagging, uploading, etc. ?
- While the main focus of this campaign is to tag the maps, everyone is very welcome to add any other tags that seem appropriate that will make the collection easier to search on Flickr, or to improve the index pages, or to upload any of the images to Commons from Flickr, or anything else that seems appropriate. The green (Book cat) templated link indicates that there is already a category on Commons with some images for a particular title. See the sections "Standard tags" and "Uploading to Commons" on the main Commons page about the collection for more information about tagging and uploading.
- If I migrate some of these images to Wikimedia Commons what tag should I add to Flickr what template do I use on commons ?
- There is an ingestion template at Commons:British_Library/Mechanical_Curator_collection#Image_descriptions. Please use this, and it will automatically set up a link-back template, and appropriate categorisation for the collection; it will also order the images from a book into page-number order, and publication-date order for a general category. (But please also then add a description, and specific categorisation for the image).
- If you are adding a number of images from a book, please also create a category for the book, with the title and author of the book in free text in the category header, plus the template {{BL1million bookcat}}; such book categories can be found linked to in green on the index pages, using the template {{HasBookCat}}, as described here.
- We haven't yet tagged images on Flickr as already migrated, but we may do this retrospectively by creating a match file from the images on commons and updating Flickr.
- Why are we using Flickr anyway ?
- At present there is very little metadata for the raw images placed by the Library on Flickr, to identify where or what they are about, or even what kind of image they are. All there is, in machine-readable form, is the name of the book and the page number. Whilst we could upload a million undifferentiated images to Commons, it was felt that this would swamp the community's capacity to sort them and descibe them and make them usefully findable -- instead just leaving Commons with a mass of unsorted, undescribed, unsearchable images to cope with.
- So, instead, specific images have been being uploaded to Commons by individual users (see here), or images a book at a time, that the uploaders have then described and categorised and given intelligible names to, so that they properly fit into Commons's image ecosystem.
- However it should be possible to upload all the maps in the collection en-masse, once they have been identified and georeferenced, using the georeferencing information to automatically place them in the right places in Commons's category system.
- How do I rotate images that are on their sides?
- The images were uploaded to Flickr just as they were scanned on the page -- so some are the right way up, and some are on their side. If you add the Flickr tags rotatec or rotatecc the images can then be rotated clockwise or counter-clockwise. These are normally picked up by the British Library within about 24 hours, and rotated versions uploaded. Copies of images that have been uploaded to Commons will usually have been rotated and appropriately cropped (see here), and may have been adjusted for colour or greyscale.
- Is there any way to see the text associated with a particular image ?
- Some of the Flickr pages include a link to "Open the page in the British Library's item viewer", but this essentially never works. The Library may in future make the fully scanned books also available through the Internet Archive, but at present this is still only under consideration.
- For the present, to see the text associated with an image a PDF of the whole book must be downloaded. For some images this is possible through a "Download PDF for this book" link in the text description on the image's Flickr page.
- For other images, there should be a link to the British Library's integrated catalogue in the description. Clicking on this, and then clicking on the "related resources" tab in red on the right of the catalogue entry, and then looking for "Refine my results" in the sidebar that comes up on the left and clicking "Online" should bring up a list of copies available as PDFs. Clicking on the title of one of the copies on this list then brings up a screen, with a link to a copy that can be downloaded.
- More questions
- For any other questions, please start a new section on the talk page of this page, and we will try to answer them.
Launch event: October 31: "Gotta Tag 'Em All!" - a digital maps Halloween tag-a-thon[edit]
- Date: Friday October 31st
- Place: British Library
- Booking: via Eventbrite
0930 - 1000 Tour of Maps area / Maps Reading Room -- Kimberly Kowal - Lead Curator, Digital Mapping
A special tour of the Maps area at the British Library - both behind-the-scenes and the reading room was available for anyone interested in printed maps, digitisation, and collection activities. The library has some surplus large-format printouts of sundry maps to choose from and take away with you.
Please meet in the front lobby at 0930 for your tour, and let us know if you are attending when you book for the event.
1000 - 1030 Coffee and Registration -- Foyle Suite, Centre of Conservation, British Library (plan)
1030 - 1030 Introduction and welcome -- Mahendra Mahey, Manager of the British Library Labs Project, British Library
1035 - 1045 Digital mapping at the British Library -- Kimberly Kowal, Lead Curator Digital Mapping, British Library
1045 - 1055 The story of the mechanical curator collection and Flickr Commons -- Ben O’Steen, Technical Lead, British Library Labs, British Library
1055 - 1105 Tagging the Mechanical Curator Collection - what you need to do! -- James Heald
1105 - 1300 Tagging!
1300 - 1400 Lunch, tagging and networking
1400 - 1410 Off the Map Competition at the British Library - 2013 and 2014 -- Mahendra Mahey on behalf of Stella Wisdom, Digital Curator, British Library
1410 - 1420 OpenStreetMap - more than a map! -- Harry Wood, OpenStreetMap
1420 - 1430 Open Data Kyoto in Practice - organising a successful OSM mapping party and Wikipedia editathon -- Machi Takahashi, Advanced Telecommunications Research Institute, Kyoto
1430 - 1620 More tagging!
1500 Working coffee break with some spooktacular treats
1520 - 1530 Wikimedia in the UK; and Wikimedia Commons -- Jonathan Cardy, Wikimedia UK
1620 Round up and summary
1630 Finish
(people are free to keep working till 1800 if they wish or attend the optional events below)
Optional events after the tag-a-thon[edit]
- 1700 - 1800 Terror and Wonder - The Gothic Imagination exhibition
- Cost: free to attendees of the tag-a-thon
Attendees may be interested in visiting the Terror and Wonder - The Gothic Imagination exhibition. ("Perversely enlightening" - The Week, quoting The Guardian). If you would like to go, please let us know on the booking form and we can arrange free attendance to the exhibition as a thank you for all your hard work!
- 1830 - 2230 LATE at the Library: The Sorting
- Cost: £15
Some of you may be interested in making a real Halloween night of it in London and attending your own funeral ;).
If you are, we can recommend LATE at the Library: The Sorting. The cost is £15 and tickets are limited! You will need to book separately for this event.
Project team[edit]
- WM: User:Jheald, User:Andrew Gray, User:Pigsonthewing (Andy Mabbett), User:WereSpielChequers (Jonathan Cardy), User:Chippyy (Tim Waters)
- BL: Kimberly Kowal, Ben O'Steen, Mahendra Mahey, Stella Wisdom
- OSM: Harry Wood
- Georeferencing community: Kimberly Kowal
Getting the word out[edit]
- Wikimedia London Meetup, Sunday 12 October (flyers ?)
- Email to prolific georeferencers
- WMUK donor list
- OSM mailing list
- Geotargeted watchlist announcement on Wikipedia in the UK's South-East
- Signpost, Kurier, worldwide Wikipedias (translation?)
- Tweets during the day -- e.g. BL / WMUK / WMF / OSM -- pics & progress, with link to progress page. Hashtag #bltagathon.
- (?) en-watchlist announcement worldwide to "Help finish finding all the maps in 1 million images from the British Library", with link to progress page (Fri 31 Oct eve / Sat 1 Nov / Sun 2 Nov ?)
Statistics[edit]
Currently in the index, there are
- 5885 titles indexed, representing 42.1% of images (430,978 images)
- 7384 titles in the "to do" pages, representing 33.6% of images (343,571 images)
- 10477 Fiction/Plays/Verse/Works titles, representing a further 20.9% of images in the collection (214,377 images)
Which makes 13,269 titles in the first two sections, or 330 each for a team of 40 taggers. (Or twice that for a team of 20).
The numbers also imply an average of just under 60 images per book. Most books contain rather less than this; but a few, a lot more.
This is more than we are likely to complete on day one. However, we should be able to get through a fair chunk of the list, kicking off a process that will steadily mop up the remainder. This mopping up process can of course be done by anybody in the world with an internet connection, plus a willingness to make simple wiki edits. Given some interested participants, it should be possible to complete the whole project in a reasonably short period of time.