Geograph Knowledgebase / FAQ

Keyword Search:


Why are you introducing Geographical Context?
Geograph publishes photos that illustrate the geography of Great Britain, Ireland and the Isle of Man.

For the first six years and 2 million images, Geograph contributors have been required to choose a primary geographical category for each of their images. They could either select one from a list or create a new one if what they wanted didn't exist. Some contributors limited themselves to a small number of very general categories (moorland, lake, buildings...) while others preferred to create much more detailed, often unique, categories (church (Roman Catholic) (former), artificial fish farming pool, bat hibernaculum...). The list grew to an unwieldy 9,000 categories, mostly of detailed photograph subjects. The primary geographical categories which offered a broad-brush way of organising the archive had got lost.

The system showed its limitations as the archive grew. For example, in this subject-rich photo link the short title and description and single category (Watercourse) mean that other features go unrecorded, so will not be picked up in a search. Similarly in link the category selected was Lake. A search on Lake would return over 25,000 images not very useful. The interesting detail in the description would only be found by a very specific search. In a third example link the category Artificial fish farming pool is unique, although the image would be found by a search for Fish farm. All these images would of course be seen in browsing their respective gridsquares. Photos with the minimum information will tend to be invisible to searches because searches are text-based. The solution is to enable contributors to choose multiple categories tags and one or more primary geographical categories geographical context - that describe the content of the photo.

Of course, submitter knows best. The new system asks submitters to choose at least one geographical context from a fixed list of 46 very general options. If several of them apply, all their boxes can be ticked. At the same time, detail is encouraged with free-form tags which would previously have cluttered up the category list, and which can now be put to good use in narrowing down searches. So, geographical context and tags together should do all that categories used to, but without their limitations. In the Askern example several context boxes could be ticked: Lake, Village, Open space, Leisure; possibly Mining. From these alone a picture forms in the mind. Tags would fill in details of the birds, the tree, and the vanished historical features. The more information that can be attached to a photo the greater its value to the archive.

[edited by Robin Stott]
· More information on this topic... · contributed by Rudi Winter, Jun 2011 · Edit this answer (Open for editing by anyone) · Provide an alternative answer!

Photo Contributors :: Contributing

How do I choose tags for my images?
The idea is that you specify at least one 'top' (or 'geographical context') tag. These are about 40 broad classes describing the main aspect of the subject at the time the image was taken. You can add additional top tags if you like, depending on what surrounds your main subject. The top tags are in the tabs labelled 'topography' to 'communications' in the tagging box.

The short list of 'top' tags avoids the clutter of the old categories, but it doesn't give you the flexibility to highlight detail that you feel is important. For that purpose, free-form tags are also available as an option. You can pick the most salient words from your description (the system may even suggest some of them in the 'suggestions' tab), or any others that you would like your image to be found by if someone uses the word as a search term.

In addition you can use prefixed tags for collections of special interest: A railway enthusiast might create a prefix 'locomotive:' and use it to tag the different engines in their pictures. Or someone interested in churches could use a 'denomination:' tag to indicate which particular community uses the place of worship shown. You can also use existing prefixes, e.g. 'place:' or 'near:' to indicate which town someting is in or near to. Have a look at what prefixes and tags others have already created: link . However, there is no need to restrict yourself in any way to that list. Prefixes are best used wherever it is likely that there will be other examples of your subject (church, locomotive...) with a slightly different attribute (Methodist, Diesel...).

As far as tags are concerned, chaotic growth is encouraged - the top tags are meant to counterbalance that!

Finally, there's buckets (another of the tabs in the tagging box). This is a limited list of special tags that are meant to limit searches to pictures of certain types (rather than certain subjects). For example, the 'gone' bucket is meant for images of which you know for certain that the geographical subject is either no longer there or has changed beyond recognition. If it's just an old picture but the landscape or buildings are the same, then I'd not use the historic bucket. See if any of the other buckets apply to your picture. Examples for some bucket types are in this incomplete article: link .
· More information on this topic... · contributed by Rudi Winter, Jun 2011 · Edit this answer (Open for editing by anyone) · Provide an alternative answer!

· Can't find the answer you looking for? Ask a question now! ·

Creative Commons Licence [Some Rights Reserved] the content of this page is licensed under a Creative Commons Licence.

You are not logged in login | register