Keyword Searching in the Browser

The Browser image searching tool has a powerful keyword search box.

It is fundamentally, the same as the keyword search on the main search, and supports most of the same syntax.

This article attempts to document the difference.

Note, the 'Browser index' is also used by the simplified 'Photos of' interface, the Groups By tester, and the coverage layer. (and a few other experimental tools)

Using "Keywords" Search box

This is finding images based on the words you enter. It looks for your keywords in an image's description, title, and other text fields. To get the best results, it's helpful to understand a few key features.

Default Search: Simply enter multiple words separated by spaces. The search will find images that contain all of these words, regardless of their order. The search is not case-sensitive.

Whole Words Only: The search matches whole words. For example, a search for river will not match riverside.

Stemming: The system uses stemming, so a search for bridge will also find bridges or bridging. This feature ensures a broader range of relevant results. Disable it for individual words by prefixing with =, for example =bridging

Searching for a Phrase: To search for an exact phrase, enclose your keywords in double quotes. For example, "road bridge" will find images that contain that specific phrase.

Using "OR": To find images that contain at least one of your keywords, use OR or the pipe symbol |. For instance, bridge OR bont will find images with either bridge and/or bont.

Excluding Keywords: To exclude a word from your search, place a minus sign in front of it. For example, canal -river will find images about canals but not those that also mention a river. This is useful for removing irrelevant results. Use minus symbol in front of every word want to exclude.

For more advanced queries and syntax options, you can consult previous help page. The above functionality is the same in both.

See below for the specific fields searched in the keywords index. By default all are searched.

Missing features

~ operator does not work in the browser
[...] tag syntax is only partly supported, not recommended to use, use tags:bridge instead. Or use attribute sidebar to search tags.
Field Start/End modifiers, don't work on Tags/Contexts etc (any field name ending in 's')

Field Searches

The browser index has its own list of available search fields.

Can use 'field syntax' to limit keywords to a specific field. If specify field name followed by colon, ALL keys after that (or until another specifier!), will only match specific field.

e.g. bridge title: canal liverpool will look for 'bridge' in ANY field. But 'canal' and 'liverpool' needs to be in the image title.

This is a full reference list: (some are intended for internal use, so not easy to use)

... also many of these fields are specifically to power the 'attribute sidebars'. So filtering by say "format" might be easier using the sidebar filters, but for all other fields can use the keywords box, and these names:-

title:	The main image title	title:bridge
desc:	The description text (note, does NOT include shared descriptions)	desc:the
name:	The full photographer/contributor name	name:John Smith
day:	The day the photo was taken, format YYYYMMDD	day:20120523
month:	The month and year the photo was taken, format YYYYMM	month:201204
year:	The year the photo was taken, format YYYY	year:2001
decade:	The decade the photo was taken, format 2010s (always ends in 0s)	decade:1990s
monthname:	The month the photo was taken, format April (the full english month name)	monthname:September
myriad:	The 100km square the image was taken in, as jsut the gridletter(s)	myriad:NT
hectad:	The 10km square the image was taken in, as 2 figure grid reference	hectad:SH55
gridref:	The 1km square the image was taken in, as 3 figure grid reference	gridref:TQ7040
category:	The legacy image category	category:ford
subject:	The selected Primary Photo Subject	subject:fungi
contexts:	The top level Geographical Context(s)	contexts:roads
buckets:	The selected bucket(s) the image has been added to	buckets:aerial
tags:	All other tags (not context, subject, buckets or types) note, if wanting to search by tag prefix, don't enter the :, eg to search the tag [river:Arthog], you need to enter tags: tags: "river Arthog"	tags:river
snippets:	The titles of the shared descriptions on the image	snippets:christmas
wikis:	The assigned wikimedia category for the image (few images have this)	wikis:bridge
terms:	Important Terms extracted by computer from description	terms:canal
groups:	Automatically cluster labels	groups:battle
status:	geograph OR supplemental	status:geograph
types:	image type tag (geograph, supplemental, cross grid, close look, extra, aerial, inside) (unknown type are 'supplemental')	types:extra
distance:	The distance between subject and photographer in metres, rounded to powers of two (eg 1, 4, 16, 256, 4096)	distance:1024
direction:	the view direction as decimal degrees from north 0-360 in multiples of 22.5 or 'unknown'	direction:180
format:	square,portrait,landscape,panorama	format:portrait
landcover:	Estimated landcover of the general area of the photo	landcover:woodland
place:	nearest placename	place:Birmingham
county:	county of the nearby placename (if known)	country:Essex
country:	The full country name	country:Ireland
postcode:	the Postcode 'district' (part before space) or 'outcode'. Note not all images are marked with a postcode, generally only in or near actual settlements. Includes UK postcodes as well as Eircodes for Ireland	postcode:EC4A
larger:	has a larger upload. (eg "larger:800" means has at least a 800px version, also 1024, 1600 and 3000 available)	larger:1024
user:	the user-id of the user who uploaded the image	user:96
text:	This is a special field, that searches the title, description, subject, category and "other" tags (in particular NOT the shared descriptions, nor the context/type tags, and contributor name)	text:mushroom

The (newer) browser and the main search, have very different fields. It has a number of fields that the main search does not.

---