It’s with some pleasure that I’m able to announce a little Xmas present from the technology team at Ingenta to the academic and RDF communities: another batch of RSS feeds from IngentaConnect.
To be precise: in excess of 20,000 new RSS feeds containing the latest table of contents data for the academic journals that are still being actively loaded into our databases. Like our friends at Nature, our feeds are available as RSS 1.0, with Dublin Core and Prism metadata at the item level.
Read on for the technical details, including a seasonal(?!) “Easter Egg”.
The feeds are clearly advertised from the journal home pages, using the well-known orange RSS icon. Auto-discovery links are also included to make one-click subscriptions possible. If you look at a journal home page, e.g. the Journal of Consciousness Studies (JCS), you’ll also see a link to the “archival” feeds that we’ve been publishing for a few months now: an RSS feed of the most recent issues of that journal. Again, the feeds are RSS 1.0 + Dublin Core + PRISM, and publishing them at both levels allows us to support varied uses of the data, e.g. import in library systems as well as current awareness applications. I know Richard Cameron has been waiting to add our TOC feeds to CiteULike, so I’ll be interested to see how people use them.
If the homepage doesn’t sport an RSS icon then its quite likely that we’re no longer hosting the most recent content of the journal. This happens some times for contractual reasons, but also because journals do fold from time to time (or become merged with other titles). Not all publishers have yet bought into the merits of metadata syndication in this form, and so in one or two places I’ve been cautious and blocked some titles. This is also the only reason why abstracts are not yet present in the feeds; but this is something I am pursuing further.
Because of the way the feeds are constructed, there is actually an RSS view of every issue on our site, it’s just that the advertised feeds use a symbolic pointer to the latest issue. Continuing the above example, you’ll see that the JCS feed is at:
However if you take an issue level URL such as:
you can extract RSS by simply tweaking the URL to point to
api.ingentaconnect.com and adding a
format=rss parameter, like so:
api.ingentaconnect.com domain is the location from which we’re going to be building out our public web services. Seems easier to manage that way.
With an RDF/RSS view of any of our issues available so easily this release constitutes more than just an bunch of RSS feeds: it actually exposes the majority of our database to the web as RDF. Hence the “Xmas present” to the RDF community. High on my TODO list for next year is to make this RDF view more explicit by adding a plain RDF view of the site (basically the same metadata shorn of the RSS trappings). FOAF support is another avenue I’m actively exploring.
The “Easter Egg” I added to the site is, ahem, an OPML export of the top-level browse function. Specifically the alphabetic, subject area, and title keyword search functionality. This will allow users to quickly subscribe to whole sections of the catalogue.
For example, if one browses in the Nursing section of the site and does a “View Source” you’ll spot an OPML “auto-discovery” link. The use of quotes there is to indicate that, as ever, there seems to be little standardisation in this area of OPML usage. The OPML link is as follows:
i.e. again change the sub-domain to “api” and add a
format attribute, this time requesting an OPML view. The page size indicator is a hack to request all items in the category. If its omitted you’ll just get the first page worth of results, sorted by journal name.
Expect more features in this area over the coming months, including feeds of new journals, and RDF Blogroll support; I’ve had a “Planet” aggregator running against a slice of our feeds for a few weeks now without any problems.
Lastly (and maybe this should have been first), if you spot any problems them please report them to me at firstname.lastname@example.org. I’ll endeavour to turn fixes round as fast as possible, just please allow us some time given the season, holiday periods, etc. I’m aware of one niggle already and I’m certain there are others.
Enjoy, and happy holidays.