Z666.7.L364 (www.jenniferlang.net)

• Carnegie Mellon Libraries:Library Arcade
  (tags: libraries information_literacy library)

As part of the Nineteenth National African American Read In and Black History Month, the Black Caucus of the American Library Association-New Jersey Chapter and the Princeton Public Library will be co-sponsoring a reading and book signing by veteran journalist Dominic Carter.

Mr. Carter will read from his memoir No Momma’s Boy at the Princeton Public Library on Sunday, February 3rd at 2:00 p.m. Copies of the book will be available for sale at the library, and after the reading Mr. Carter will autograph copies.

From a childhood of abuse at the hands of his mentally unstable mother to a meaningful career attending social events at the White House and interviewing Nelson Mandela, No Momma’s Boy recounts Carter’s tremendous highs and lows and how he survived to become a successful journalist.

A fixture at NY1 News since the channel launched in 1992, Dominic Carter hosts the nightly political show “Inside City Hall,” during which he has handled some of that station’s most challenging assignments and interviewed every major political player in the state. Carter also appears regularly on CNN and MSNBC.

Mr. Carter received a bachelor’s degree in journalism from SUNY Cortland and a master’s from the Newhouse Graduate School at Syracuse University before starting his career in radio.

The African American Read-In, sponsored by the Black Caucus of the National Council of Teachers of English, urges citizens nationwide to make literacy a significant part of Black History Month by hosting and coordinating read-ins in their communities.

All Princeton Public Library programs are free and open to the public. If programs require registration, preference is given to library cardholders. The physically challenged should contact the library at (609) 924-9529, 48 hours before any program with questions about special accommodations.

The library is in the Sands Library Building at 65 Witherspoon St. in Princeton Borough. Convenient parking is available on neighboring streets and in the borough-operated Spring Street Garage, which is adjacent to the library. For more information about library programs and services, call (609) 924-9529 or visit www.princetonlibrary.org.

The ALCTS Electronic Resources Interest Group met on Saturday, January 12, 2008, from 10:30am-12:00pm. The program was well attended (about 115 people signed in, but the actual headcount was closer to 200), especially given the number of other programs occurring in the same time slot.

The topic of session was making e-resources more manageable and other issues related to records, in particular issues related to e-book aggregations and databases.

Our first speaker was Nicole Pelsinsky, who recently joined Serials Solutions as a product manager focused on the 360 MARC Updates service. Prior to joining Serials Solutions, Nicole worked at Microsoft Corporation for seven years. Some of her previous projects include the multifaceted search experience for the Microsoft Library Global Market Research Search Development, a glossary lookup tool, an image search system, and search usability testing. Before joining Microsoft, Nicole worked with digital asset metadata for Getty Images. In more traditional library roles, her work included positions at Johns Hopkins University and the Centers for Disease Control.

Nicole began by noting that she is able to bring a unique perspective to her work at Serials Solutions based her prior work on the customer side—issues from the library perspective, something “near and dear to her heart.” The problem she found with e-resources of any type, especially when she was working at Microsoft, is that there was a lack of discoverability. People couldn’t find the resources unless they were represented in the OPAC. If the resources weren’t in the OPAC and the library didn’t provide some other way for users to access the resources (e.g., facilitating browsing through the library portal), people were mystified about why it was that they weren’t having a good user experience and being connected with information. The other thing that was an issue was that with all the changes in the data with the subscriptions that they had, they did not have enough staff and resources to keep up with all of those changes. They were constantly scurrying around trying to figure out ways to discover broken links, etc. Also, many times their customers would discover problems before staff would, so there was a lot of retroactive cleanup. Furthermore, a new e-book subscription might mean thousands of MARC records to be loaded into the OPAC, so just the sheer volume would “terrify” her IT people, since they were the ones who were supporting the library and loading the records in the OPAC.

At this point, Nicole assured the audience that there are solutions for these issues, but first she gave us the provider perspective based on her experiences at Serials Solutions. Now that she is at Serials Solutions, Nicole has more of an understanding of the issues they face as a content provider:

• there are varying levels of completeness of records and the standards are all over the place, so many times the information needs to be manipulated in order to conform to some sort of standard
• the metadata for e-books is not always reliable – sometimes it may not exist at all, sometimes it may not be complete
• keeping up with changes and growth of subscriptions is problematic – titles come and go in the packages and links change often
• providers need to do a better job of defining standards they want from publishers, in terms of content and data
• the PCC has gone a long way in developing standards and Serials Solutions has reviewed those standards, but the many times there is very useful data beyond just what conforms to a MARC standard and it would be very useful for them to have “forward thought” about how they could use that data and make it available to libraries

So what can Serials Solutions do for libraries?

• improve the publisher provided data and, more importantly, they can improve the relationships between them as a provider of content to libraries and publishers
• providers can improve relationships between themselves and the publishers and the librarians – eliminate the lack of good support (e.g., whom to speak to about a problem)
• supplying standardized frame work for delivering data and delivering it to libraries is really crucial – tools are being developed and they are working populating their repository with content other than e-serials and e-journals; are working on how to put it in a framework that is usable for libraries
• they will be offering customization of records for libraries’ local practices – they want to make sure they conform to libraries’ local practices; the Serials Solutions records won’t look any different
• they provide coverage of a variety of electronic resources from different publishers and deliver it to library in one package; instead of trying to go and get e-book information for several publishers, perhaps e-resource database from another place, SS will bring all of that information together and deliver it to the library in one package
• behind the scenes – make regular updates to the data; have a title authority control system in place and are using it for e-serials and e-journals and will be expanding use to e-books and other types of content; they provide holdings data; and they have brief record generation to provide a broader coverage of a libraries’ collection

Another point Nicole emphasized was finding a solution to the question “what is the most crucial issue for them as a content provider?” For Serials Solutions, the answer is really trying to get better data from publishers, which comes down to building relationships and trying to make it mutually beneficial for Serials Solutions, the publishers, and the libraries.

Serials Solutions is starting with its journal collections for their KnowledgeWorks certification program and will be expanding it to e-books and other types of content. The requirements that publishers have to adhere to are that they supply title list and a dedicated contact person. In return, the publisher will get a dedicated Serials Solutions contact person so that troubleshooting will be easier, etc. Publishers will also get access to the Serials Solutions interface so they can see what libraries are seeing when they are looking at their information; they can see firsthand the problems that a library might be experiencing. There is also an element of mutual publicity that goes along with this program – a good way to tie all of the pieces together and share all of the pieces of information. The end result of certification is having a collaborative environment with publishers. Serials Solutions can do lots of data manipulation in house, but it isn’t scalable in the long term. So, getting the data from the source puts them in a better position to provide better service to libraries.

Click here to view Nicole’s PowerPoint slides.

Our second speaker was Timothy Savage from OCLC. Tim received a bachelor of music degree from Oberlin Conservatory and an MLIS degree from the University of Pittsburgh. He has been at OCLC for nine years working in various capacities—first as a music cataloger in TechPro and later as a project manager. Currently he serves as a product manager in Cataloging and Metadata Services where he manages vendor partnerships, particularly those involving electronic content. Chief among his duties are managing and automating the cataloging activities for numerous partners in a variety of formats including e-books, e-audiobooks, e-music, and e-video.

Tim spoke about OCLC’s activities in terms of creating records for electronic resources, particularly their efforts to automate record creation.

Dramatis personæ

Tim likes to draw a distinction between publishers and aggregators. He defines publishers as “those who publish.” They own the rights to their content. Aggregators, on the other hand, host the published content and license the content from the publishers. Oftentimes, aggregators can also be publishers and vice versa.

According to Tim, there is a great deal of electronic content “out there” so it only makes sense to try introduce automation to try to cope with the cataloging load. Automation in cataloging isn’t new, however. Devices such as constant data, macros and “newing” records have been used for years. OCLC Contract Services has actually been working with aggregators to create records for econtent for nearly five years, so this really isn’t that new, but there is always room to improve upon their processes.

Tim pointed out that the LC Working Group’s report highlighted the fact that there is a great deal of redundancy in our cataloging workflows in terms of duplicate data entry taking place, and perhaps more surprisingly, duplicate subject analysis. Many publishers assign BISAC headings, which is a subject vocabulary created by the Book Industry’s Study Group, to their metadata records. There is a great opportunity to try to leverage some of that subject work into the assignment of LC subject headings and even into the derivation of call numbers.

But as we move forward with automation, Tim points out that we (as catalogers) shouldn’t lose sight of our core values:

• performing subject analysis and classification
• determining the relevancy of access points
• performing authority control

But as we move into the future, Tim suggests that there are some new values we might want to consider, namely, the value that can be gained from using vendor-supplied metadata to try to automate the way a lot of the descriptive tasks and the descriptive parts of cataloging are done. Tim admitted having spoken to many librarians who are understandably concerned that introducing data from these sources and interest in automation compromises quality or “dumbs down” the catalog, but he doesn’t agree. Tim thinks we really need to focus on improving how we do things so they can be done with greater efficiency so that we can have more things cataloged and, therefore, more things discoverable. By automating a lot of the “descriptive minutia,” we can repurpose our intellectual efforts to those tasks that we deem the most critical for resource discovery. He finds it interesting that the LC Working Group’s report says that more research is needed to determine those elements of the bibliographic record that most heavily drive resource discovery because he suspects that if this research were done, the results would point back to the three core values he mentioned earlier.

So what is OCLC doing?

Building on the success of the past five years of partnering with aggregators, OCLC is looking to “take it to the next level” by not only receiving metadata from aggregators but by going “further upstream” directly to the publishers themselves. Tim described OCLC’s “next generation cataloging effort”—they are receiving metadata directly from publishing community in an XML standard called ONIX (Online Information Exchange), a standard that is widely used across the publishing industry. As titles are being made ready for publication, publishers will often pass metadata on to places like Amazon.com or Barnes & Noble, and that data automatically gets uploaded onto those sites and the items described by that metadata become available on that vendor’s website for sale. Since this metadata is already available, it makes sense for OCLC to tap into it as well. OCLC is using automated processes to convert the ONIX data into MARC, and in so doing, make the records more robust and complete in accordance with our current cataloging standards.

At this point, Tim showed the audience a sample ONIX record and pointed out what data could be used, such as ISBN, title, author, imprint data, and even subject terms, which can be leveraged by matching these terms to controlled vocabularies.

OCLC has established a “template-based cloning process” in which they take the best parts of the ONIX record and combine these with the best parts of data that is already in WorldCat. Tim describes the process as a sequence of constant data layers being applied one on top of the other. Using the ONIX metadata received from the provider, OCLC searches and mines WorldCat to find records to clone of the same title that are perhaps in a different form (e.g., a print version of an e-book or a CD of an e-music resource). If there is not a match, OCLC follows a born-digital process. If there is a match to clone, OCLC deletes unnecessary fields such as 533 fields from cloned microform records. Then they apply data from a format template so that, for example, all e-music records created will have the necessary 006 and 007 fields, there will be descriptive notes that state that it is streaming music or downloadable music, etc. For further consistency, OCLC applies constant data to describe the aggregator. Currently with e-books, they include a 533 field as well as added entries identifying the aggregator to make the records more consistent. Next, title specific data from the ONIX record is added, including the eISBN and the URL for the title so they can seamlessly link from the catalog to the e-resource. Once all of the automated pieces have been “stacked on top of each other,” OCLC performs a quality review to ensure that records are complete—primarily checking for the presence of call numbers, subject headings, and URLs. They also perform authority control.

Tim displayed a flowchart that illustrates the process:

start with vendor ONIX record…

• includes print ISBN which is valuable search key
• includes title information to help verify that they have a correct record to clone
• eISBN and URL is batched to be added to new record

once there is a match…

• harvest from print record: fixed field, call numbers, authorized headings
• notes fields, especially for e-music
• subject headings are retained

in the template system…

• fields are added: 006, 007
• make sure GMDs are applied consistently
• notes describing online characteristics are added
• add heading for the aggregator
• add text to the 856 subfield $z describing the URL

this all results in…

• a complete record containing all of the required data elements

“Easier said than done.”

Tim pointed out that there are challenges that come along the way, one of which is the multiplicity of approaches to cataloging that proves problematic. For example, different standards and tags are used depending on whether the item is considered to be a reproduction or if it is considered to be born digital. Tim made the point that from the sound recordings perspective, which is his background, considering electronic items as reproductions and cataloging them as such is kind of like saying a CD is a reproduction of an LP. You wouldn’t use an LP with an LP’s 300 field and then have a 533 field saying “compact disc”—this can be misleading to users. Another inconsistency is that even if we were to make that distinction between reproductions vs. born digital, Tim doesn’t see it applied to every format. It seems to be very common with e-books but he doesn’t see it with e-videos or e-music; these last two formats seem to be considered born digital even if they are re-releases of content that was previously issued on another medium.
There has also been variance in terms of MARC tagging—elements in one record appear in one field and the same kind of information may appear in a different field in a different record. Also, when searching for e-records to use, Tim points out the match may be to a record that has a URL that is specific to one particular institution, which won’t be of value to another institution. Another issue is that the ISBNs are no longer unique. Tim has found that some publishers will assign a new ISBN to an electronic version of a title while others will use the print ISBN; obviously this complicates not only the searching, but especially the automated database searching. Even if there is an eISBN it is often used by multiple aggregators so that it still ceases to be unique.

Multivolume titles are also problematic depending on how the title is presented on an aggregator’s website and described in that aggregator’s metadata, and also on how a previous version of that same title was cataloged in WorldCat. There might be a single record with a single URL pointing to a “landing” page from which one can link to all volumes, or there is a single record with multiple URLs, or there might be multiple records, one for each volume. This situation complicates the automated searching and matching process, and makes it difficult to clone records and document each URL properly.

Aggregator-neutral records

Tim raised the question of whether or not aggregator-neutral records can be used effectively for electronic monographs. He thinks they can, if all electronic resources were described as born digital. He made the point that a single record could become very bloated with 533s for each aggregator and that this would be confusing for users. Certainly there would be multiple URLs added to an aggregator-neutral record, but if we consistently identify the URLs for each vendor by perhaps using an 856 subfield $3, Tim believes that the management of the URLs could be automated (in fact, OCLC is in the process of developing a tool to manage the 856 fields such that deletion of unnecessary 856 fields could be automated). Aggregator neutral records would make searching more efficient by having so many fewer hits than we have today, and also he thinks automated editing could be more efficient by having all of the data elements consistently applied with the same tagging, instead of like today, where there may be a 533 field in a reproduction record with “mode of access” notes and in a born digital record, the same information is in a 538 field. These sorts of discrepancies make automated processing difficult.

In summary, Tim thinks that core cataloger values like subject analysis, authority control, and evaluation of access points are here to stay, despite increased automation. He does believe, however, that we need to actively need to seek out and embrace anything we can to improve our workflows so that we can get these things cataloged more quickly so that they are discoverable through the OPAC. Tim also thinks the cataloging rules need to be more consistent and simplified in order for automation to achieve its best potential.

Click here to view Tim’s PowerPoint slides.

Peter Fletcher, our third speaker, joined the UCLA Cataloging and Metadata Center in July 2007 as the Cyrillic scripts metadata and cataloging specialist. He has participated in PCC since 1998 and has been a CONSER cataloger since 2002. Prior to UCLA, Peter worked at Tulane University as a library instructor, a bibliographer for Germanic and Slavic materials, and a serials and e-resources cataloging librarian. Peter also chaired the PCC Task Force on Provider Neutral Records for Electronic Integrating Resources.

The title of Peter’s presentation was “Provider Neutral Record for Remote Access Electronic Integrating Resources.” He began by reviewing the AACR2 definition of an integrating resource: “A bibliographic resource that is added to or changed by means of updates that do not remain discrete and are integrated into the whole; can be finite or continuing. Examples include updating Web sites.”

In the current cataloging world, each vendor version of an online integrating resource receives a separate cataloging record. For example, the online version of Art Index, there are several catalog records, each representing a different provider or vendor such as OCLC.

Peter gave a little background of the policy: it is based on serials aggregator-neutral policy for online remote access serials that CONSER implemented in 2003. Currently, there are many records in WorldCat for online integrating resources such as the MLA Bibliography which is available from various providers such as CSA, Ebsco, OCLC FirstSearch, and ProQuest. There are records for each provider, and even though CSA has bought ProQuest, records describing both versions are still in WorldCat. WorldCat is full of duplicate records for online integrating resources, so what we need is a provider neutral record that is based on the original publisher’s content.

The idea of a provider-neutral record for online integrating resources was first discussed at a CONSER meeting in 2005, and the response was generally positive. It was suggested that catalogers on discussion lists be contacted to get a better sense of the response and it, too, was positive. After this, a task force was formed to make recommendations and it issued its final report in September 2007. At the PCC Policy Committee meeting in November, the recommendations in the report were approved. Currently, the BIBCO integrating resources cataloging manual is being revised to include the provider neutral instructions and the revised draft should be available by the end of January 2008.

Some key aspects of the policy:

• preferred source for cataloging will be the publishers website whenever possible (provider version is the second choice)
• ISSN will be assigned to represent the original version
• the digital provider other than the original publisher will only be recorded in the record to indicate which version the cataloging was based on, and an access point will be created only for the original publisher or society
• summary information in the record should be applicable to all versions

The goal is to have one authoritative record, and this policy should make selecting a record in OCLC for use much easier by reducing the number of duplicate records. The record will be simpler and more adaptable.

Click here to view Peter’s PowerPoint slides.

After the presentations were over, a brief question and answer period followed:

Mary Mastraccio informed the group that MARCIVE has been working with ERIC to make MARC records available for ERIC documents. She also reiterated what Nicole Pelsinsky said in her presentation: that collaboration is important – libraries need to be involved to see how the MARC records that the vendors are supplying work in their local catalogs.

Greta de Groat (Stanford University) shared her concerns about the availability of records for titles in packages such as SourceOECD (is anyone working on records for these?). Also, she wondered about records for free resources such as working papers from the United Nations because the large institutions may not want to pay for records for resources that they aren’t purchasing. In addition, there are projects that individual institutions are doing such as the Missouri Botanical Garden Library for which there are records in OCLC but they aren’t offered as a set. There are a lot of freely available packages out there but how are we being notified of new additions and record availability? If we could get the records from a vendor or OCLC, it would be very helpful.

Tim Savage responded that libraries could let OCLC know what packages we need/want records for. New WorldCat Collection Sets can be created—we just need to let OCLC know. Also, if records are being created for local projects but not being added to WorldCat, then perhaps the institutions could contribute the records to a union catalog (WorldCat).

Nicole Pelsinsky mentioned that Serials Solutions has sent out a survey about types of information institutions need from publishers and they are working on summarizing the results.

Someone asked for more information about OCLC’s link m