BHC2492 The Passenger Liner ‘Queen Mary’ Arriving at Southampton, 27 March 1936
(Repro ID: BHC2492 © National Maritime Museum, Greenwich, London)
Last week, Mike Ellis posted about RSS feeds in museums. Specifically, how useful it would be if search results were available as newsfeeds. After a bit of tinkering around, and a fair amount of swearing at catalogue descriptions written in the Windows extended character set, I’ve now set up feeds for the National Maritime Museum collections. If your web browser supports RSS, then you should be able to find a feed on almost any page that generates a list of catalogue records. Here are some examples – a search for ‘tower bridge’; objects from the Atlantic Worlds galleries; paintings and drawings by Charles Pears; photographs of the Aquitania; relics found at Erebus Bay. The collections search is also available via OpenSearch, which I’ve tested in Firefox and IE7. If your browser supports OpenSearch, then ‘NMM Collections’ should be available as a search engine from any page under http://www.nmm.ac.uk/collections/. This is really cool, as it opens up our collections to be used by any application that can consume RSS. I’ve also extended the news feed items with catalogue metadata using dc:coverage (publication info), dcterms:spatial (geographical coverage) and dcterms:temporal (date made), which opens up the possibility of plotting the objects on maps or timelines.
superb! we’re hopefully getting a google mini appliance at work soon (our department is fronting some seed money to get it for two years, as our IT department is ridiculously shit at this sort of thing), and i’m itching to work on customising a nice, usable, sexy search front end and semantic results pages, including options to get results in various formats like RSS/Atom.
one minute quibble: can your RSS feed output be changed, so that the title of the feed reflects the search terms that were entered? e.g. “Results for: ‘tower bridge’ – National Maritime Museum Collections Online” or similar?
Cheers Patrick. We were looking at the bug with the feed titles this afternoon. Although the logic that runs a search is seperated out from the view templates, allowing us to pipe results to either HTML or RSS templates, the logic that generates the page heading is old, old code embedded in the HTML view template. Should be fixed tomorrow (fingers crossed) so that the feed title matches the heading on the corresponding web page.
The feeds have individual titles and descriptions now!
marvelous…now i’m definitely itching to get my toy…
Nice work (and nice blog – added to my feedreader).
We’ve got a Google Mini running. So give me a buzz if you want to compare notes. I’m sure we’re not using ours to its full advantage yet.
I found a RSS-to-KML converter. Here’s a link to a KML file generated from the first twenty records in our charts collection:
http://eatyourgreens.org.uk/testapps/rssToGeoRSS.kml
and the feed displayed in google maps.
Is that Dublin Core I spy? Yuck.
P.S. We have a Google search appliance at my work and people are going crazy suggesting sites to index. How exciting.
Not just any old Dublin Core, but Qualified Dublin Core, taken from the PNDS Application Profile.
Hey Jim,
thanks for the link and also the info – looking really cool. Frankie and Mia and I were talking about using the GSA / GM for this kind of stuff. I have access to a full GSA at work for my own evil doings, so was going to work up ways in which it could deliver an API-like set of search results, probably OpenSearchy or whatever. Anyway, I’ll watch this thread with interest and punt anything that might be useful to you.
Meanwhile (plug) don’t forget to add stuff to http://www.mashedmuseum.org.uk as and when you invent it 🙂
Cheers
Mike
Hi Jim.
I came across you blog through mashedupmuseum. I thought you might find my latest post interesting. I actually used the NMM OpenSearch feeds for my experiments several times.
Cheers
Giv
Did anyone ever end up writing code to do good things with a google mini? Because now I’ve got one to play with, and we’re not doing much with it…
Not that I’m aware of, Mia, although Mike Ellis wrote something ages ago about playing with a Google Search Appliance. I would be very interested in hearing from anyone who has done anything with the Maritime Museum feeds. There’s also a YQL interface to this data.
Europeana did use the feeds to harvest records for their index.