Cataloging Futures

The relevance of quality cataloging

2012-05-28T10:50:37-04:00

Software developer from Harvard's Library Innovation Laboratory, Paul Deschner, writes about the importance of quality cataloging for the development of new library applications.

One of the primary challenges in this work is getting data describing books and periodicals (catalog records) to relate to data from non-library sources, such as data about book talks on YouTube or to NPR broadcasts of author interviews or to archival collections. It’s all about connections in the data. The barer the data, the less describedit is, the more it falls flat.

On the bibliographic side, every new Library of Congress subject heading a cataloger adds to a record creates a rich set of connective possibilities downstream for people like me. Likewise, every uniform title entry inserted into a record allows us to show users of our software another edition of a given work in the context of all its editions — a crucial feature for any discovery service in the library materials space.

No software can create these connections if the underlying data hasn’t been carefully composed into richly structured records, based on solid analysis and comprehensive description. The difference is like that between reading a newspaper consisting of headlines only and reading one which also has accompanying articles. It is dramatic.

I hope in moving forward that we don’t lose sight of the importance of this kind of quality analysis and description.

He goes on to describe the essential contribution that catalogers make to the development team:

At the Law Library, the catalogers are a few hallway steps away, and are as crucial to my being able to create smart software as anyone on my development team. I’ve spent countless hours, regularly throughout the years, with my cataloger colleagues exploring the complexities of MARC data structures, uniform title rulesets, authority record uses, holdings data locations, and much much more. Having them as a co-located resource has been crucial to my being able to get my software written.

h/t @john_overholt

Cataloging Futures: change of focus a misstep

2012-05-26T08:48:56-04:00

Back in January (almost five months ago) I thought I was going to take this blog in a new direction. But, you know what, I made a mistake.

I realized that the future of cataloging and metadata in libraries is a topic I'm still involved with, beguiled by, and enmeshed. My work in digital collections has me going back to my roots in cataloging frequently, practically daily. And MARC is now just MARCXML.

So, I'm back to the original focus of this blog and hopefully can make a worthy contribution to the ongoing discussion.

Now time to catch up.....

Cataloging Futures takes a new direction

2012-01-05T11:20:27-05:00

Those of you who've been reading this blog know that it's focus has been the future of cataloging and metadata in libraries. I started the blog in April 2007 when this was a "hot topic" and also a personal passion. At the time I was the head cataloger of a small cataloging department with a strong background in cataloging and RDA. When I first started Cataloging Futures there weren't many catalogers blogging (that's really changed!), so it seemed to fill a much-needed niche.

For the past four years, I've been the metadata librarian on our gradually growing digital collections team. And last month I started a new job: XML Database Administrator. Our team works with a production mindset in a agile environment. We work in focused two-week sprints. Both work and learning my new job don't afford me the luxury of keep up with the trends and issues of cataloging and metadata with the intensity that blogging requires.

So, while the blog is still called "Cataloging Futures," the focus with now be primarily using XQuery (and MarkLogic Server) for digital collections development in libraries. I hope that some of my current readers stay on board, but totally understand in the busy work-a-day world of cataloging that you might have other things to focus on. I am truly grateful for all the Cataloging Futures readers who've encouraged and challenged me over the last several years. Thank you!

Interesting post: CodeYear Catalogers

2012-01-04T08:57:38-05:00

If you're interested in learning programming, check out Shana L. McDanold's blog post: CodeYear Catalogers.

CodeAcademy is offering an excellent opportunity for anyone interested in getting started with coding: Code Year. Each week they'll be emailing out to anyone that signs up an interactive lesson on coding. Code Academy offers some fantastic interactive courses, so you learn in a supportive environment.

Naturally this has piqued the interest of a number of catalogers that have been interested in learning to code or want to brush up or learn new code languages, myself included. Catalogers work with massive amounts of curated bibliographic data, and being able to manipulate it in new and different ways and in ever increasing amounts is key as we move forward into the bibliographic future and the world of linked data and the semantic web.

XQuery for library metadata: no. 4

2011-09-28T10:50:52-04:00

Here's a query that catalogers will love! It leverages Library of Congress Classification numbers from MARCXML records to create subject facets. We used this for a discovery portal for digitized theology books. (This application is not live yet, but I'll add the link when it is.)

We got the idea for this query from Kurt Cagle's example which makes an XQuery typeswitch expression more like a traditional switch. Each LC class is turned into a temporary element to get the typeswitch to work. The end result is a new class element with a value assigned for the subject facet.

Also, This query was refactored and turned into a library module. I'd be glad to share it (thought it was too long for a blog post).

xquery version "1.0-ml";

(: 2/12/11 Query for mapping LCC to subject categories using typeswitch expression. This updated 
version includes all of the LCC, not just the B class. :)

declare default element namespace "http://www.catalogingfutures.com/ia";
declare namespace m = "http://www.loc.gov/MARC21/slim";

for $meta in xdmp:directory("/xml/")/doc/metadata
let $oldNode := $meta/class
let $marcMeta :=$meta/marc/m:record
let $lcClass := ($marcMeta/m:datafield[@tag=("050", "055", "090")]/m:subfield[@code="a"])[1]
let $bsClass := for $classNum in $lcClass
                where fn:starts-with($lcClass, "BS")
                return if (fn:number(fn:substring($lcClass, 3, 4)) < 701)
                       then "BS1"
                       else if (fn:number(fn:substring($lcClass, 3, 4)) < 1901)
                            then "BS2"
                            else if (fn:number(fn:substring($lcClass, 3, 4)) > 1900)
                                 then "BS3"
                                 else ()
let $bvClass := for $classNum in $lcClass
                where fn:starts-with($lcClass, "BV")
                return if (fn:number(fn:substring($lcClass, 3, 4)) < 590)
                       then "BV1"
                       else if (fn:number(fn:substring($lcClass, 3, 4)) < 2000)
                            then "BV2"
                            else if (fn:number(fn:substring($lcClass, 3, 4)) < 3750)
                                 then "BV3"
                                 else if (fn:number(fn:substring($lcClass, 3, 4)) < 4200)
                                      then "BV4"
                                      else if (fn:number(fn:substring($lcClass, 3, 4)) < 4485)
                                           then "BV5"
                                           else if (fn:number(fn:substring($lcClass, 3, 4)) 
<= 5099)
                                                then "BV6"
                                                else ()
let $bxClass := for $classNum in $lcClass
                where fn:starts-with($lcClass, "BX")
                return if (fn:number(fn:substring($lcClass, 3, 4)) < 100)
                       then "BX1"
                       else if (fn:number(fn:substring($lcClass, 3, 4)) < 800)
                            then "BX2"
                            else if (fn:number(fn:substring($lcClass, 3, 4)) < 4800)
                                 then "BX3"
                                 else if (fn:number(fn:substring($lcClass, 3, 4)) <= 9999)
                                      then "BX4"
                                      else ()
let $bClass :=  if (fn:matches($lcClass, "^B\D\d"))
                   then fn:substring($lcClass, 1, 2)
                   else ()
let $azClass := if ($lcClass)
                then fn:substring($lcClass, 1, 1)
                else ()
let $newClass := ($bsClass, $bvClass, $bxClass, $bClass, $azClass, "Unclassified")[1]
where $oldNode = "Unclassified"
return for $class in $newClass
       let $c := element { $class }{}
       return let $newNode := typeswitch ($c)
                               case $c as element (B) return <class>Philosophy</class>
                               case $c as element (BC) return <class>Logic</class>
                               case $c as element (BD) return <class>Philosophy</class>
                               case $c as element (BF) return <class>Psychology</class>  
                               case $c as element (BH) return <class>Aesthetics</class>
                               case $c as element (BJ) return <class>Ethics</class>
                               case $c as element (BL) return <class>Religion</class>
                               case $c as element (BM) return <class>Judaism</class>
                               case $c as element (BP) return <class>Islam</class>
                               case $c as element (BQ) return<class>Buddhism</class>
                               case $c as element (BR) return <class>Church History</class>
                               case $c as element (BS1) return <class>Bibles</class>
                               case $c as element (BS2) return <class>Old Testament</class>
                               case $c as element (BS3) return <class>New Testament</class>  
                               case $c as element (BT) return <class>Theology</class>
                               case $c as element (BV1) return <class>Worship</class>
                               case $c as element (BV2) return <class>Ecclesiology</class>
                               case $c as element (BV3) return <class>Missions</class>
                               case $c as element (BV4) return <class>Practical Theology</class>
                               case $c as element (BV5) return <class>Preaching</class>
                               case $c as element (BV6) return <class>Practical Theology</class>
                               case $c as element (BX1) return <class>Ecumenism</class>
                               case $c as element (BX2) return <class>Eastern churches</class>         
                               case $c as element (BX3) return <class>Catholic Church</class>          
                               case $c as element (BX4) return <class>Protestantism</class>
                               case $c as element (A) return <class>General Works</class>
                               case $c as element (C) return <class>History</class>
                               case $c as element (D) return <class>History</class>
                               case $c as element (E) return <class>History</class>
                               case $c as element (F) return <class>History</class>  
                               case $c as element (G) return <class>Geography and Anthropology</class>
                               case $c as element (H) return <class>Social Sciences</class>
                               case $c as element (J) return <class>Political Science</class>
                               case $c as element (K) return <class>Law</class>
                               case $c as element (L) return <class>Education</class>
                               case $c as element (M) return <class>Music</class>
                               case $c as element (N) return <class>Fine Arts</class>
                               case $c as element (P) return <class>Language and Literature</class>
                               case $c as element (Q) return <class>Science</class>
                               case $c as element (R) return <class>Medicine</class>  
                               case $c as element (S) return <class>Agriculture</class>
                               case $c as element (T) return <class>Technology</class>
                               case $c as element (U) return <class>Military Science</class>
                               case $c as element (V) return <class>Military Science</class>
                               case $c as element (Z) return <class>Bibliography</class>
                               default $c return <class>Unclassified</class>
           return xdmp:node-replace($oldNode, $newNode)

XQuery for library metadata: no. 3

2011-09-26T11:47:21-04:00

It's been about six months since I shared some XQuery code. Most of the code I write is used to search and manipulate metadata for our digital collections. I have a lot stockpiled, so the next few posts will be some XQuery examples.

This query was write to modify METS documents for a digital archival collection. We wanted a user interface feature that would allow the user to both show or hide blank pages for each individual archival item. Those who wanted the "feel" of the orginal artifact could see the blank pages. And those just interested in the content could hide them.

We identified the blank pages, so I had the blank page file names to work with. I wanted to add a TYPE attribute in the METS structMap div elements for all the blank pages: TYPE="blank-page".

This query was the solution. I worked my way from the fileSec to the structMap matching on identifiers until I got to the correct div element that needed the added blank page attribute. The MarkLogic function xdmp:node-insert child() was used to add the attributes to the correct div element.

xquery version "1.0-ml";

(: 9/16/11 Query for adding blank page attributes to structMap of METS documents using a sequence

of file names from blank page reports :)

declare namespace mets = "http://www.loc.gov/METS/";
declare namespace xlink = "http://www.w3.org/1999/xlink";

let $doc := fn:doc("/METS/Kuyper239-241.xml")
let $file-names := ("/240-0001t", "/240-0002t", "/240-0004t", "/240-0006t", "/240-0008t")
let $images := for $file-name in $file-names 
               return fn:concat("Manuscripts/thumb/Kuyper/239-241", $file-name, ".jpg")
for $image in $images
for $file-location in $doc/mets:mets//mets:FLocat
where $image = $file-location/@xlink:href
return for $id in $file-location/parent::mets:file/@ID
       for $file-id in $doc/mets:mets/mets:structMap/mets:div/mets:div/mets:div/mets:fptr
       where $id = $file-id/@FILEID
       return xdmp:node-insert-child($file-id/parent::mets:div, attribute TYPE { "blank-page" })

Controlled vocabularies: more than just words

2011-06-29T08:28:31-04:00

I think this quote from an article titled "More Than Words" nicely describes some of the goals behind library linked data:

... Over the past decade, a bioinformatics specialization called biomedical ontology has grown up around this question. Most biomedical researchers are familiar with standard terminologies, such as Medical Subject Headings. These are sometimes called ontologies, but true ontologies are more than just controlled terms. They capture, in a logical, systematic way, what scientists regard as the basic truths about a topic. Like equations in physics or axioms in mathematics, they can even be the basis for computational models. When connected to databases, scientific papers, and software applications, ontologies "help cope with the ever-growing, chaotic accumulation of text and facts" in biomedical and translational research. They do this by making data sharing, retrieval, and validation easier, says Stefan Schulz, a professor at the Medical University of Graz in Austria.

h/t Linked Data and URI:s for Enterprises

ALA 2011: Some RDA news

2011-06-27T08:25:08-04:00

The RDA Toolkit blog has an interesting post on RDA news coming out of the American Library Association Annual Conference. Here's a snippet:

It was a busy day at the ALA Conference for RDA and RDA Toolkit. The U.S. Test Committee met with vendors, testers, and the public, and summarized the results of the test and the reasoning behind their recent decision. The committee stated that the "rewrite" of RDA will actually be more of a rewording, intended to simplify and clarify some of the instructions. Cataloging practice will not be affected by the rewording. We also learned that the Library of Congress will begin partial RDA cataloging in November. They will catalog across all subject areas and expect RDA records to represent about 5% of the library's cataloging output. The National Library of Medicine and the National Agricultural Library will not catalog in RDA until full implementation begins.

From catalogers to information ninjas?

2011-06-24T07:18:32-04:00

Here's a thought-provoking summary of a presentation from this year's CLA conference: Goodbye Librarians, Hello Information Ninjas! Many of the ideas address a movement away from tradtional cataloging activities. Some excerpts:

Models that define information roles and information services based on content type distinctions are becoming less relevant – more likely than not, we will begin seeing overlaps and complementary domains developing.

The great challenge of this “digital age” we find ourselves in will be ensuring we do what we do (the collection, preservation and management of information) in a way that ensures it is accessible in the future. No matter its format or content, we are really all about the information at the end of the day – and what’s it worth if we can’t access it?

So what are the biggest foundational pieces of what will be required of our new role as Information Ninja?

Figuring out the best way to manage and access resources, wherever they are located and in whatever system and format they are stored.

Moving away from traditional records file and library classification systems and making innovative use of more consumable methods of classification such as metadata and social tagging.

Focusing on establishing common classification and management practices, applicable to any information resource, across systems and formats.

Recognizing and building the necessary skills: becoming tech savvy; building knowledge of search; understanding structured and unstructured content; learning about retention and preservation requirements, just to name a few.

MARBI and the future of the MARC format?

2011-06-21T09:03:43-04:00

Yesterday on the MARC listserv, Elizabeth Plantz (National Library of Medicine) and Tina Shrader (National Agricultural Library) raised concerns as to why the Machine-Readable Bibliographic Information Committee (MARBI) has nothing on their agenda on the future of MARC for the upcoming ALA Annual Conference meetings.

Their questions for MARBI are worth reading as the Library of Congress begins its work investigating a possible transition away from the MARC format, a 40+ year old library metadata structure.