I have come up with two bookmarklets that allow you to search for an author’s works in a library catalogue from the author’s Wikipedia page in one click. A bookmarklet is a browser bookmark that does something with the page you’re looking at rather than just taking you to a web page: see the helpful Firefox guide for more information. The bookmarklets are identical, except that one searches UCL’s Explore (Primo) service, and the other searches COPAC. To try them:
Install one of the bookmarklets by dragging the link to your bookmarks toolbar:
You can rename them to something more snappy if you like. Next, go to a Wikipedia page for an author. The bookmarklets only work on Wikipedia pages with VIAF or LC Authorities links in them, but most major authors should be fine. Some examples to try:
- Steve Jones (has a VIAF link only)
- Charles Darwin (has LC link only)
- Charles Darwin (a different one: has VIAF link only, and try on COPAC)
- Charles Dickens (has both VIAF and LC links)
Why. One of the promises of linked data and BIBFRAME and all the rest of it, is that data from different sources can be linked together and work with each other. Since VIAF links were recently added to Wikipedia, I’ve wondered what could be done to take advantage of this in a practical way. The link does mean that from a Wikipedia page and its uncontrolled (or at least only consistent within Wikipedia) names, you can find out the authorised form of an author’s name. Charles Darwin (the famous one) is only called Charles Darwin in the title of his Wikipedia article. Search for that on a library catalogue and you’ll get all his works plus the stuff written by other Charles Darwins. With the VIAF data, we know that he is known in most (or a huge number of English-language catalogues) as “Darwin, Charles, 1809-1882” as opposed to the other Charles Darwin mentioned above, who is “Darwin, Charles, 1758-1778”. Although most catalogues or discovery systems don’t use linked data and non-textual identifiers, the ubiquity and uniqueness of an LC heading does almost perform a similar thing (although there are caveats galore).
Many of the caveats are in the way library systems search. Both the examples used are imperfect: the UCL one, as I’ve done it, uses a facet search on top of the bare search which while eliminating some incorrect hits (where “Steve”, “Jones”, and “1944-” appear coincidentally as author elements in a search) also misses out a few hits depending on what field he appears in in the record (this is I think a fixable glitch which I intend to get fixed); the COPAC one is author free text but I’ve tried to remedy some of the potential for false hits by putting all searches in quotes.
Improvements. These are legion, but a few sketched ideas below:
- Implement this as a browser extension. This was my original intention, so that someone could be browsing any old Wikipedia page and when they come across one with a VIAF (or other service) link at the bottom, a search link is created at the top of the page for them to click on. This could easily be extended in several ways:
- Add subject searches. Should be straightforward, although would require moar bookmarklets, relying on the PHP script offering options, or a proper browser extension
- Add more catalogues/discovery interfaces. This is again straightforward to add to the PHP script if you can figure out the web API for a search service but is subject to the same caveats as subject searches.
- Add more than just VIAF and LC Authorities. There are other links appearing at the bottom of some Wikipedia pages, most notably Worldcat. The bookmarklet itself could be easily adapted to accommodate these so avoiding a further profusion of bookmarklets, as well as more back up when services are down (VIAF went down twice while I was testing). Adding services to the PHP is a matter of knowing the structure of the RDF, which shouldn’t be too painful.
- Improve how errors are dealt with and reported, especially so the bookmarklet handles more of them and prevents the PHP script being called unnecessarily.
Feedback. I appreciate this is highly unlikely to set the world on fire, but I would be interested in any feedback or ideas of how it could be developed. Of course, please do let me know if you come across any mistakes or problems: it’s becoming almost traditional for me to get the most crucial link wrong in blog posts.