Saturday, October 2, 2010

2010/09/27-30 - moving through jello

I spent most of this week working on a submission tool for the minnows morphology project. The following e-mail describes what I'm implementing.

... Reuben Pasquini 09/30/10 1:01 PM ...
Hi Jon,

I've been looking at how to manage submissions to the morphology repository,
and I have some ideas to bounce off you.
I'd like to implement the following process.

  *. We'll setup a custom submission form that only allows a user
    to upload a single .zip file.
    The .zip file may contain an arbitrary number of .tps files,
    and the .jpg images that go with them.

  *. A user uploads the zip file, and fills out a form of metadata.

  *. The server automatically unzips the zip file,
    and breaks it down into individual items like
       http://131.204.172.126:8080/minnows/handle/123456789/10 
   , and submits the items for review after the user
   verifies that everything looks ok.

  *. The reviewer approves or rejects each individual item.     

That's the direction I'm working in now - sound ok ?

Cheers,
Reuben

BTW - If you want to allow excel spread-sheeting in addition to .tps files, 
    then send me instructions on how the data in a spread sheet
    maps to the data in a .tps file, and I'll code that in too.

I felt like I was moving through jello this week - writing code and doing little things, but nothing released to users. I did contribute (see below) to an e-mail exchange about testing article-level search in our vufind catalog.


... Reuben Pasquini 09/30/10 10:36 AM ...
Hi Nancy!

Thanks for taking the time to write down your thoughts - you have
great ideas.  I'll help stir the pot a little - see if we can get her to boil!

*. Your argument about wanting students
           "
             to think more critically about
             where and what they're searching for information
            "
      made me think that I actually want the opposite.
      I don't want to think at all - just give me what I want.

*. Your point about the apparent overlap between the vufind, voyager-2, 
    and EBSCO efforts is well taken.
    It makes sense to have vufind, voyager-2, and WAG
    under one umbrella.
    I don't think it's a big problem though.
    Clint or Tony are in every meeting, so they keep things connected.
    I think they like meetings.

* Google Scholar:
      http://scholar.google.com/intl/en/scholar/about.html 
   "
Google Scholar aims to rank documents the way researchers do, weighing the full text of each document, where it was published, who it was written by, as well as how often and how recently it has been cited in other scholarly literature.
   "
    Sounds awesome.

*. "Are we using our technical and human resources correctly?"
    Are you having doubts only now ?
    You probably still leave cookies out for Santa ...


Cheers,
Reuben

... "Nancy Noe"  9/29/2010 8:56 PM ...
Hi Marliese

Working tonight at the Village.  I've been able to play around with the
Villanova site a bit, but haven't given it the time or consideration it
deserves.  These are my initial impressions.  In the spirit of
full-disclosure, I should probably say that my comments come from a very
specific information literacy philosophy.

One of the concerns I have with this kind of searching is the same issue
I have with federated searching.  While I understand the 'google-like'
feel of such a search, I want students to think more critically about
where and what they're searching for information.  There was a recent
discussion on the  instruction list serv - are people teaching federated
searching/discovery services? While a couple of people said 'yes', the
majority replied in the negative.   Again, it comes back to helping
students to learn to move away from the 'everything possible' approach,
to one that asks them to consider what disciple/subject specific
resources might provide them with higher quality academic/scholarly
material.  I want them to think critically about their information need
before they start typing in poorly constructed, broad-based keyword
searches, especially when looking for articles.   For instance, I taught
an honors ENGL1127 today, working on literacy issues.  Within the EBSCO
suite, I was able to have them select ASP and three education related
databases.  That's the kind of thing I'm going to teach in class. 
That's where I'm going to fall on the issue.

Technically, I found that in the searches I conducted, there were a lot
of 'clicks' one had to go through before you got to the actual article. 
In two instances, there was a link to full-text that failed (and that
was prior to asking for log-in information.)

Vendors are creating systems that in a way compete with Google Scholar,
yet I suppose there's a 'cost' for these systems.  Perhaps I don't fully
understand, but won't they be charging us for the privilege of using
metadata when we already pay for access to the info through our database
subscriptions?  Does this give us permanent access to the articles? 
Faculty tell me that they now use Google Scholar as the starting point
for their research.   Many schools have a link to Google Sciholar off
their library homepage.  Would this serve us just as well?   I may not
be as informed as I need to be on cost and ownership issues.

Should the VuFind cmte investigate?  Maybe.  We're still working on
trying to make the catalog work (advanced searching?) and trying to
determine if we should recommend as primary.  To be honest, the past
couple of weeks I've found myself reverting back to the 'classic'
catalog to answer journal title questions.  I haven't settled on
questions I have about VuFind as a catalog.   I'm not saying that we
shouldn't be exploring new systems/new technologies.  We should.  Is
this the right cmte to take this forward?  We have the new voyager cmte,
and the EDS cmte, and the vufind cmte?  Are we using our technical and
human resources correctly?  Is there a way to better optimize time and
talents, especially since it seems that all of these are converging.

So, those are my two cents this evening.  Again, I'll miss the meeting
tomorrow.

Thanks

Nancy

... "Marliese Thomas"  09/29/10 2:52 PM ...
Hi everyone, 

Here's some background about the opportunity I mentioned at our last
meeting.

We've been talking for a while about taking VuFind to the next level by
including article-level metadata in it.  Indeed, at our August 9th
meeting, Bonnie volunteered to contact vendors about getting a batch of
article-level metadata that we could run through Vufind in order to see
how this might work.

There have been a couple of new developments since that meeting. 
First, Bonnie, Marcia, Aaron, and I met with Jane Burke and Mary Miller
of Serials Solutions a couple of weeks ago.  In the course of that
meeting, Jane told us that Serials Solutions has developed a programming
interface that allows other discovery services, including VuFind, to
search Summon and display results from it.  This functionality allows
VuFind to do article-level searches in Summon without the host library
having to ingest, store, and index large article-level metadata sets
provided by vendors.

To see how the VuFind-Summon combination works in practice, go to
Villanova's VuFind catalog at:

https://library.villanova.edu/Find/Search/Home 

Do a search.  In addition to the familiar-looking VuFind results,
you'll see a box in the upper right-hand corner called "Top results from
Articles & more", with links.  "Articles & more" = Summon.  One example:
https://library.villanova.edu/Find/Search/Results?lookfor=cat&type=&search=catalog+AllFields&submit=Find 


The day after we met with the Summon folks, EBSCO announced that EBSCO
Discovery Service has a similar functionality.  Since we're a full-level
development partner for EDS, we can test this feature with VuFind at no
cost to us.  (Of course, we'd have to buy EDS, or Summon, or a similar
product, in order to offer article-level content to our users as part of
our regular services.) Since then, I have also heard that Ex Libris is
offering this same functionality for their discovery product, Primo
Central. Obviously, this is becoming a common and viable way for
libraries to leverage use of their software contracts.

I believe it would be worth exploring how VuFind works with a
full-featured commercial discovery service like EDS.  In fact, Bonnie
gave Aaron and me the go-ahead to take a shot at getting a VuFind-EDS
test installation working in time for my presentation this weekend at
LITA.  I'd be interested in hearing your thoughts on these developments
at our meeting tomorrow, or when you have had a chance to explore the
Villanova catalog.

Thanks,

Marliese



Marliese Thomas
Database Enhancement Librarian
RBD Library
Auburn University
 
231 Mell Street
Auburn, AL 36849
 
334.844.8171
mst0001@auburn.edu 

Finally, Thursday was Penny's last day. Enjoy your retirement, Penny!

No comments:

Post a Comment