headers
Karen Coyle | 1 Oct 08:06

Re: Google Booksearch Data API: Another blow to library metadata

Tim Spalding wrote:
> The Google API has what Amazon lacks--coverage of
> out-of-print books.
Actually, Amazon does have o-o-p books, but they are the ones submitted 
by their bookselling partners and are often of very poor quality. We 
crawled Amazon for the Open Library project and got a number of these. 
We'll replace them with better records from libraries should we get them.

kc
Permalink | Reply | 0 comments |
headers
Karen Coyle | 1 Oct 08:22

Re: Library Technologies and Library School (was Commercial Vendors and Open Source Software)

As for selling v. free -- librarians want something that will be there 
for the foreseeable future -- they don't want their system breaking 
randomly because someone decided to close a site and go meditate on a 
mountain. Unfortunately, many 'free' services tend to be ephemeral. With 
$$, you sign a contract, and that contract has penalties for 
non-performance, so the company is pretty much constrained to keep 
going. Not that companies have never failed libraries, but they have 
almost certainly done so less than 'free' services.

Eventually, we'll have free services that have enough of a track record 
to be trusted, but as we know, libraries work on a long time frame -- 
not years, but at least decades.

kc

Tim Spalding wrote:
> My big hope is for the OpenLibrary project, but, while I don't follow
> Open Library as closely as I was once, I get the feeling that it's not
> winning—it's not succeeding in breaking the library data out, getting
> enough non-library or user data to mount a challenge that way, or
> involving enough people in the library community. I'm doing a lot of
> publishing talks now, and I haven't met anyone who's even heard of it.
> Open sourcing book data seems like a complete no-brainer, and yet...
>
> Similarly, although LibraryThing has basically let go of a million
> covers and series data as good as anyone else's, and it's going
> nowhere. What works for us—and it's working like hell—are the things
> we sell as ready-made services (tags, recommendations, soon reviews).
> I would bet you anything that, if we had *sold* the covers, they'd be
> used more. Seriously. Maybe we could sell them and then—psych—not send
(Continue reading)

Permalink | Reply | 2 comments |
headers
Bernhard Eversberg | 1 Oct 08:40
Picon
Picon
Favicon

Re: Library Technologies and Library School (was Commercial Vendors and Open Source Software)

Karen Coyle wrote:
> 
> Eventually, we'll have free services that have enough of a track record 
> to be trusted, but as we know, libraries work on a long time frame -- 
> not years, but at least decades.
> 

Yes. That means, the best scenario would be to have a not too small
number of libraries who commit themselves to at least some development
work, and to have some longer-term staff for that purpose, not just
projects that run at most two years and are then shut down.
These libraries would band together as a developers' community
which could of course be open for global participation.
This is not impossible. VuFind seems to be setting an example, and
there is sufficient stuff out there so that one need not start
from scratch.
We have had a similar model running here for over 25 years, and
hundreds of special libraries are successfully using the software.
Those who can afford it can also call in freelance supporters instead
of providing the manpower themselves.
(Our LCSH browser, for example, is based on it. This way, it could
be rigged up in a matter of days. Some config work, some PHP, was all.)

B.Eversberg
Permalink | Reply | 0 comments |
headers
Weinheimer Jim | 1 Oct 08:54

Re: Library Technologies and Library School (was Commercial Vendors and Open Source Software)

Karen Coyle wrote:

> Eventually, we'll have free services that have enough of a track record
> to be trusted, but as we know, libraries work on a long time frame --
> not years, but at least decades.

This may be true, but I don't think that libraries can afford to wait decades (which in internet terms is
eons). Already, and I'm not kidding, many of my students cannot even imagine what it was like to do any kind
of research before Google (What? No copy and paste?! People had to *write things down*??!!!) For them,
it's like people were living in the Stone Age running from mammoths and saber-tooth tigers.

So, maybe while libraries are waiting around for systems to trust, our users will have moved on long before
and forgotten us completely. It's a quandry for libraries, I admit, but I don't think we can afford to
ignore the vast majority of materials that our users are demanding today just because they don't fit into
our usual framework.

Many of these resources disappear, as you say, but this is where we could work closely with the Internet
Archive, which has an immense number of sites stored. If something isn't being archived, we could
probably get it on a fast track somehow. If that's not suitable, perhaps we could actually build something ourselves?

The main thing is: we need to show that we are relevant today.

Jim Weinheimer
Permalink | Reply | 0 comments |
headers
Stephens, Owen | 1 Oct 10:53
Picon

Re: Google Booksearch Data API: Another blow to library metadata

Somewhere in this thread (I think) there was discussion of around XML
and bib data, and MODS was mentioned as a better encoding of bib data
into XML than MARCXML (which some felt wasn't a particularly difficult
feat). Anyway, I thought that this might be of interest
http://copac.ac.uk/development-blog/2008/09/persistent-identifiers-for-c
opac-records/ - COPAC is a union catalogue for many of the largest UK HE
Research Libraries, and they are doing some interesting development work
now - in this particular case they are exposing records in MODS - 'tho
you have to know the record ID

Owen Stephens
Assistant Director: eStrategy and Information Resources
Central Library
Imperial College London
South Kensington Campus
London
SW7 2AZ

t: +44 (0)20 7594 8829
e: o.stephens <at> imperial.ac.uk
Permalink | Reply | 4 comments |
headers
Jan Szczepanski | 1 Oct 16:15
Picon
Picon
Favicon

English will not only do, but do better

I disagree with Alexander Johannesen and Bernhard Eversberg about the 
language
question and just want to point to the fact that in Europe all languages 
are equal and
I enclose an exampel:

+ TOWARDS MULTILINGUAL “SEARCH & RESULT”
One of The European Library’s objectives is to provide information in 
all partners native languages; this means having the portal interface 
translated in all full members’ languages. It also implies a more 
ambitious target that is to extend the multilingual capabilities of the 
portal; in other words, provide a language understanding of queries and 
return the appropriate search results in the same language. EDLproject 
Work Package 2 aims at integrating the outcomes of multilingual access 
research into The European Library portal.
The European Library language policy follows similar guidelines to 
Europa, the portal site of The European Union: “As far as possible, the 
aim is to provide the public with the information they are looking for 
in their own language.” This means that the language interface of the 
HOME and COLLECTIONS pages are translated in all Full Participants’ 
languages. The latest update of the portal interface involves the 
integration of the Russian translation provided by the Russian State 
Library.
On the other hand, EDLProject Work Package 2 (WP2), Extending the 
multilingual capacity of the network, is developing the European Library 
network's localisation and multilingual capabilities by improving access 
for end-users. This is generated through multi-language interfaces and 
advanced search mechanisms in a standardised way. EDLProject WP2 is led 
by the National Library of Slovenia with the input of the Swiss National 
Library. Since EDLproject is coming to an end, we have asked Genevieve
(Continue reading)

Permalink | Reply | 627 comments |
headers
Jonathan Rochkind | 1 Oct 16:56
Favicon

Re: English will not only do, but do better

Yes, for users, whether in Europe or North America or anywhere else, we 
need to support multi-lingual displays. Perhaps even for staff users 
such as catalogers, sure.

That doesn't mean our machine data files need to be multi-lingual. 

Jonathan

Jan Szczepanski wrote:
> I disagree with Alexander Johannesen and Bernhard Eversberg about the 
> language
> question and just want to point to the fact that in Europe all 
> languages are equal and
> I enclose an exampel:
>
>
>
> + TOWARDS MULTILINGUAL “SEARCH & RESULT”
> One of The European Library’s objectives is to provide information in 
> all partners native languages; this means having the portal interface 
> translated in all full members’ languages. It also implies a more 
> ambitious target that is to extend the multilingual capabilities of 
> the portal; in other words, provide a language understanding of 
> queries and return the appropriate search results in the same 
> language. EDLproject Work Package 2 aims at integrating the outcomes 
> of multilingual access research into The European Library portal.
> The European Library language policy follows similar guidelines to 
> Europa, the portal site of The European Union: “As far as possible, 
> the aim is to provide the public with the information they are looking 
> for in their own language.” This means that the language interface of
(Continue reading)

Permalink | Reply | 0 comments |
headers
Alexander Johannesen | 1 Oct 17:42
Picon

Re: English will not only do, but do better

On Wed, Oct 1, 2008 at 16:15, Jan Szczepanski <jan.szczepanski <at> ub.gu.se> wrote:
> I disagree with Alexander Johannesen and Bernhard Eversberg about the
> language question and just want to point to the fact that in Europe all languages
> are equal and I enclose an exampel [...]

I am myself of both Norwegian, Swedish and Czech heritage, speak 4
languages, married to an Aussie with bi-lingual kids, and I've got
relatives across the globe all speaking different languages. I know
the importance of providing multi-lingual interfaces and backends to
people, don't you worry. :)

I think perhaps the definition of "the language question" is at fault
here; I'm purely speaking of an XML vocabulary in this case, not the
content of that data model. Heck, why do you think I'm such a Topic
Maps fan? :)

<topic id="34">
   <name type="#english">Hi there</name>
   <name type="#english-aussie">G'day</name>
   <name type="#norwegian">Heisan</name>
   <name type="#swedish">Tjenare</name>
</topic>

Notice here that the XML vocabulary is in English, while the content
itself is in its respective language with the appropriate attributes
to denote language.

Alex
--

-- 
---------------------------------------------------------------------------
(Continue reading)

Permalink | Reply | 627 comments |
headers
Dobbs, Aaron | 1 Oct 18:30
Favicon

Re: English will not only do, but do better

Not to open the MARCXML is [good/bad] argument again...
(Leaving out discussions on AACRx punctuation and capitalization practices, etc., too)

Is a rethink of MARCXML, or whatever, similar to:

<topic id="245">
   <name type="#english">Title</name>
   <name type="#spanish">Título</name>
   <name type="#german">Titel</name>
</topic>

Different that what you're saying?
Is the "245" moniker fairly standard?
Is it too abstract for mapping?

-Aaron
:-)'

-----Original Message-----
From: Next generation catalogs for libraries [mailto:NGC4LIB <at> LISTSERV.ND.EDU] On Behalf Of Alexander Johannesen
Sent: Wednesday, October 01, 2008 11:42 AM
To: NGC4LIB <at> LISTSERV.ND.EDU
Subject: Re: [NGC4LIB] English will not only do, but do better

On Wed, Oct 1, 2008 at 16:15, Jan Szczepanski <jan.szczepanski <at> ub.gu.se> wrote:
> I disagree with Alexander Johannesen and Bernhard Eversberg about the
> language question and just want to point to the fact that in Europe all languages
> are equal and I enclose an exampel [...]

I am myself of both Norwegian, Swedish and Czech heritage, speak 4
(Continue reading)

Permalink | Reply | 1 comment |
headers
Alexander Johannesen | 1 Oct 18:45
Picon

Re: English will not only do, but do better

On Wed, Oct 1, 2008 at 18:30, Dobbs, Aaron <AWDobbs <at> ship.edu> wrote:
> <topic id="245">
>   <name type="#english">Title</name>
>   <name type="#spanish">Título</name>
>   <name type="#german">Titel</name>
> </topic>
>
> Different that what you're saying?

Yes, very; in Topic Maps, everything is a <topic> (and if you wanna go
completely abstract, nutty and datamodally-correct, so is is
associations, types, and occurrences; they're just special <topics>,
or shorthands, if you will), and the meaning of each topic is handled
through a proxy of other topics (through associations), types (topics
used as type) and occurrences (a special topic type). This is a meta
model in which you create your more specific model.

> Is the "245" moniker fairly standard?

The id attribute is completely up to any closed system to make up as
they go along, and is not something you can rely on. But, there's
hope. Names are pretty much ajour with titles and so forth, but the
topic at hand might be tht book you've got in your hand, Here's pretty
much the same;

<topic id="344573587345897">
   <name>Some title</name>
   <name type="#title-responsibility">some responsibility</name>
   <name type="#title-form">letter</name>
</topic>
(Continue reading)

Permalink | Reply | 0 comments |

Gmane