Interlinking of collections: the quest continues

After an excellent talk today about LibraryThing by LibraryThing's Tim, I got enthused to see how LibraryThing stacks up against other libraries for having matches in it's authority control system for entities we (the NZETC) care about.
The answer is averagely.
For copies of printed books less than a hundred years old (or reprinted in the last hundred years), and their authors, LibraryThing seems to do every well. These are the books likely to be in active circulation in personal libraries, so it stands to reason that these would be well covered.
I tried half a dozen books from our Nineteenth-Century Novels Collection, and most were missing, Erewhon, of course, was well represented. LibraryThing doesn't have the "Treaty of Waitangi" (a set of manuscripts) but it does have "Facsimiles of the Treaty of Waitangi." It's not clear to me whether these would be merged under their cataloguing rules.
Coverage of non-core bibliographic entities was lacking. Places get a little odd. Sydney is "http://www.librarything.com/place/Sydney,%20New%20South%20Wales,%20Australia" but Wellington is "http://www.librarything.com/place/Wellington" and Anzac Cove appears to be is missing altogether. This doesn't seem like a sane authority control system for places, as far as I can see. People who are the subjects rather than the authors of books didn't come out so well. I couldn't find Abel Janszoon Tasman, Pōtatau Te Wherowhero or Charles Frederick Goldie, all of which are near and dear to our hearts.

Here is the spreadsheet of how different web-enabled systems map entities we care about.

Correction: It seems that the correct URL for Wellington is http://www.librarything.com/place/Wellington,%20New%20Zealand which brings sanity back.

eBook readers need OpenURL resolvers

Everyone's talking about the next generation of eBook readers having larger reading area, more battery life and more readable screen. I'd give up all of those, however, for an eBook reader that had an internal OpenURL resolver.

OpenURL is the nifty protocol that libraries use to find the closest copy of a electronic resources and direct patrons to copies that the library might have already licensed from commercial parties. It's all about finding the version of a resource that is most accessible to the user, dynamically.

Say I've loaded 500 eBooks into my eBook reader: a couple of encyclopedias and dictionaries; a stack of books I was meant to read in school but only skimmed and have been meaning to get back to; current block-busters; guidebooks to the half-dozen countries I'm planning on visiting over the next couple of years; classics I've always meant to read (Tolstoy, Chaucer, Cervantes, Plato, Descartes, Nietzsche); and local writers (Baxter, Duff, Ihimaera, Hulme, ...). My eBooks by Nietzsche are going to refer to books by Descartes and Plato; my eBooks by Descartes are going to refer to books by Plato; my encyclopaedias are going to refer to pretty much everything; most of the works in translation are going to contain terms which I'm going to need help with (help which theencyclopedias and dictionaries can provide).

Ask yourself, though, whether you'd want to flick between works on the current generation of readers---very painful, since these devices are not designed for efficient navigation between eBooks, but linear reading of them. You can't follow links between them, of course, because on current systems links must point either with the same eBook or out on to the internet---pointing to other eBooks on the same device is verboten. OpenURL can solve this by catching those URLs and making them point to local copies of works (and thus available for free even when the internet is unavailable) where possible while still retaining their

Until eBook readers have a mechanism like this eBooks will be at most a replacement only for paperback novels---not personal libraries.

Thoughts on koha

The Koha community is currently undergoing a spasm, with a company apparently forking the code.
As a result a bunch of people are looking at where the community should go from here and how it should be led. In particular the idea of a not-for-profit foundation has been floated and is to be discussed at a meeting early tomorrow morning .
My thoughts on this issue are pretty simple:

• A not-for-profit is a fabulous idea
• Reusing one of the existing software not-for-profit (Apache, Software in the Public Interest, etc) introduces a layer of non-library complexity. Libraries are have a long history with consortia, but tend to very much flock together with their own kind, I can see them being leary of a non-library entity.
• A clear description of a forward-looking plan written in plain language that everyone can understand is vital to communicate the vision of the community, particularly to those currently on the fringes

Data and data modelling and underlying assumptions

I feel that there was a huge disconnect between some groups of participants at #opengovt (http://groups.google.co.nz/group/nzopengovtbarcamp) in Wellington last weekend. This is my attempt to illuminate the gaps.

The gaps were about data and data modelling and underlying assumptions that the way one person / group / institution viewed a kind of data was the same as the way others viewed it.

This gap is probably most pronounced in geo-location.

There's a whole bunch of very bright people doing wonderful mashups in geo-location using a put-points-on-a-map model. Typically using google maps (or one of a small number of competitors) they give insights into all manner of things by throwing points onto maps, street views, etc, etc. It's a relatively new field and every time I look they seem to have a whizzy new toy. Whizzy thing of the day for me was http://groups.google.com/group/digitalnz/browse_thread/thread/b5b0c96ce08ca441 . Unfortunately the very success of the 'data as points' model encourages the view that location is a lat / long pair and the important metric is the number of significant digits in the lat / long.

In the GLAM (Galleries, Libraries, Archives and Museums) sector, we have a tradition of using thesauri such as the Getty Thesaurus of Geographic Names. Take all look at the entry for The Wellington region:http://www.getty.edu/vow/TGNFullDisplay?find=wellington&place=&nation=New+Zealand&prev_page=1&english=Y&subjectid=7000512

Yes, if has a lat and a long (with laughable precision), but the lat and long are arguably the least important information on the page. There's a faceted hierarchy, synonyms, linked references and type data. Te Papa have just moved to Getty for place names in their new site (http://collections.tepapa.govt.nz/) and frankly, I'm jealous. They paid a few thousand dollars for a licence to thesaurus and it's a joy to use.

The idea of #opengovt is predicated on institutions and individuals speaking the same languages, being able to communicate effectively, and this is clearly a case where we're not. Learning to speak each others languages seems like it's going to be key to this whole venture.

As something of a worked example, here's something that I'm working on at the moment. It's a page from The Manual of the New Zealand Flora by Thomas Frederick Cheeseman, a core text in New Zealand botany, see http://www.nzetc.org/tm/scholarly/tei-CheManu-t1-body1-d22-d5.html The text is live on our website, but it's not yet fully marked up. I've chosen it because it illustrates two separate kinds of languages and their disparities.

What are the geographic locations on that page?

* Nelson-Mountains flanking the Clarence Valley
* Marlborough—Kaikoura Mountains
* Canterbury—Kowai River
* Canterbury—Coleridge Pass
* Otago—Mount St. Bathan's

The qualifier "2000–5000 ft" (which I believe is an elevation range at which these flourish) applies across these. Clearly we're going to struggle to represent these with a finite number of lat/long points, no matter how accurate. In all likelihood, I'll not actually mark up these locations, since the because no one's working with complex locations, the cost benifit isn't within sight of being worth it.

Te Papa and the NZETC have a small-scale binomial name exercise underway, and for that I'll be scripting the extraction of the following names from that page:

* Notospartium carmichœliœ (synonym Notospartium carmichaeliae)
* Notospartium torulosum

There were a bunch of folks at the #opengovt barcamp who're involved in the "New Zealand Organisms Register" (http://www.nzor.org.nz/) project. As I understand it, they want me to expose the following names from that page:

* Notospartium carmichœliœ, Hook. f.
* Notospartium torulosum, Hook. f.

Of course the name the public want is:

* New Zealand Pink Broom
* ? (Notospartium torulosum appears not to have a common name)

Note that none of these taxonomic names actually appear in full on the page...


Yes is, clearly, an area where the best can be the good and visa versa, but the good needs to at least be aware of the best.

Learning XSLT 2.0 Part 1; Finding Names

We mark up a lot of names, so one of the first things I decided to do was to build an XSLT stylesheet that takes a list of names and tags those names when they occur in a separate XSLT file. To make things easier and clearer, I've ignored little things like namespaces, conformant TEI, etc, etc.

First up, the list of names, these are multi-word names. Notice the simple structure, this could easily be built from a comma seperated list or similar:

<?xml version="1.0" encoding="UTF-8"?>
<names>
<name>Papaver argemone</name>
<name>Papaver dubium</name>
<name>Papaver Rhceas</name>
<name>Zanthoxylum novæ-zealandiæ</name>
</names>

Next, some sample text:

<?xml version="1.0" encoding="UTF-8"?>
<doc>
<p> There are several names Papaver argemone in this document Papaver argemone</p>
<p> Some of them are the same as others (Papaver Rhceas Papaver rhceas P. rhceas)</p>
<p> Non ASCII characters shouldn't cause a problem in names like Zanthoxylum novæ-zealandiæ AKA Zanthoxylum novae-zealandiae</p>
</doc>

Finally the stylesheet. It consists of three parts: the regexp variable that builds a regexp from the names in the file; a default template for everything but text(); and a template for text()s that applies the rexexp.

<?xml version="1.0"?>
<xsl:stylesheet version="2.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform" >

<!-- build a regexp of the names -->
<xsl:variable name="regexp">
<xsl:value-of select="concat('(',string-join(document('name-list.xml')//name/text(), '|'), ')')"/>
</xsl:variable>

<!-- generic copy-everything-except-texts template -->
<xsl:template match="@*|*|processing-instruction()|comment()">
<xsl:copy>
<xsl:apply-templates select="@*|*|processing-instruction()|comment()|text()"/>
</xsl:copy>
</xsl:template>

<!-- Look for binomal names in appreviated form where the genus name is in the immediately preceeding head -->
<xsl:template match="text()">
<xsl:analyze-string select="." regex="{$regexp}">
<xsl:matching-substring>
<name type="taxonomic" subtype="matched">
<xsl:value-of select="regex-group(1)"/>
</name>
</xsl:matching-substring>
<xsl:non-matching-substring>
<xsl:value-of select="."/>
</xsl:non-matching-substring>
</xsl:analyze-string>
</xsl:template>

</xsl:stylesheet>

The output looks like:

<?xml version="1.0" encoding="UTF-8"?><doc>
<p> There are several names <name type="taxonomic" subtype="matched">Papaver argemone</name> in this document <name type="taxonomic" subtype="matched">Papaver argemone</name></p>
<p> Some of them are the same as others (<name type="taxonomic" subtype="matched">Papaver Rhceas</name> Papaver rhceas P. rhceas)</p>
<p> Non ASCII characters shouldn't cause a problem in names like <name type="taxonomic" subtype="matched">Zanthoxylum novæ-zealandiæ</name> AKA Zanthoxylum novae-zealandiae</p>
</doc>

As you may notice, I've not yet worked out the best way to handle the 'æ'

Legal Māori Archive

Now that the Legal Māori Archive is live, I thought I'd highlight a couple of my favourite texts from the corpus.

The first is a great example of reinforcing cultural confusion. "The Laws of England, Compiled and translated into the Māori language" by judge Francis Dart Fenton is a bi-lingual compendium of the laws of England, but extraordinarily uses bible quotes as examples.

The second example is actaully a collection of texts, the works of Rev. Henry Hanson Turton, who compiled thousands of pages of land deeds and associated documents into six volumes. I can see these seeing a lot of use by Treaty researchers.

Why card-based records aren't good enough

Card catalogs have a long tradition in librarianship, dating back, I'm told, to the book stock-take in the French revolution. Librarians understand card catalogs in a deep way that comes from generations of librarians having used them as a core professional tool all their professional lives. Librarians understand card catalogs in ways that I, as a computer scientist, never will. I still recall on one of my first visits to a university library, I asked a librarian where I might find books by a particular author, they found the work for me arguably as fast as I can now find works with the new wizzy electronic catalog.

It is natural, when faced with something new, to understand it in terms of what we already know and already understand. Unfortunately, understanding the new by analogy to the old can lead to form of the old being assumed in the new. It was true that when libraries digitized their card catalogs in the 1970s and 1980s, they were more or less exactly digital versions of the card catalog predecessors, because their content was limited to old data from the cards and new data from cataloging processes (which were unchanged from the card catalog era) and because librarians and users had come to equate a library catalog with a card catalog---it was what they expected.

MARC is a perfect example of this kind of thing. As a data format to directly replace a card catalog of printed books, it can hardly be faulted.

Unfortunately, digital metadata has capabilities undreamt of at the time of the French revolution, and card catalogs and MARC do a poor job of handling these capabilities.

A whole range of people have come up with criticisms of MARC that involve materials and methodologies not routinely held in libraries at the time of the French revolution (digital journal subscriptions and music, for example), but I view these as postdating card catalogs and thus the criticism as unfair.

So what was held in libraries in 1789 that MARC struggle with? Here's a list:

• Systematically linking discussion of particular works with instances of those works
• Systematically linking discussion of particular instances with those instances ("Was person X the transcriber of manuscript Y?")
• Handling ambiguity ("This play may have been written by Shakespeare. It might also have been a later forgery by Francis Bacon, Christopher Marlowe or Edward de Vere")

All of these relate to core questions which have been studed in libraries for centuries. They're well understood issues, which changed little in the hundred years until the invention of the computer (which is when all the usually-cited issues with MARC began).

The real question is why we're still expecting an approach that didn't solve the problems two hundred years ago to solve our problems now? Computers are not magic in this area they just seem to be helping us do the wrong things faster, more reliably and for larger collections.

We need a new approach to bibliographic metadata, one which is not ontologically bound to little slips of paper. There are a whole range of different alternatives out there (including a bevy of RDF vocabularies), but I've yet to run into one which both allowed clear representation of existing data (because lets face it, I'm not going to re-enter worldcat, and neither are you, not in our lifetimes) and admitting non-card-based metadata as first class elements.

</rant>