1. Welcome, introductions and announcements. Arno Kastner, Chair)

   After the introductions the Chair (Arno Kastner, NYU) reported on last night’s dinner with OCLC senior management.  Glenn Patton of OCLC will provide a summary report which Arno will distribute to Big Heads.

    

2. Election for Vice-Chair

   Bob Wolven  (Columbia) will take over as Chair at the end of this conference. Karen Calhoun (Cornell) was elected Vice-Chair, Chair Elect.

    

3. Big Heads membership

               Every 3 years the Big Heads review their membership based on the ARL criteria index ranking; however Stanford is no longer a member of ARL. In addition, if recent trends continue, there will soon be 3 Canadian members according to the current criteria, whose libraries are rarely represented at ALA.   Should we therefore revise the criteria and/or look at attendance as a factor?  The Chair asked for volunteers to review these 2 questions, suggesting that we might want to look behind the data that we get from ARL.  Judi Nadler (Chicago) agreed that it is necessary to look at what is behind ARL statistics.  Bob Wolven (Columbia) said it might be better to appoint a committee in 2 years when we have more data though there was feeling that it would be useful to have conversation in the intervening period.  It was agreed to postpone formal study of the membership criteria.

    

    

4. CONSER summit (Summit on Serials in the Digital Environment, March 18-19, 2004, Alexandria, VA)

   Bob Wolven (Columbia) said lots of information about the summit is available on the CONSER website (http://lcweb.loc.gov/acq/conser/summit.html).  The summit will be discussed at the CONSER (Cooperative Online Serials)-at-Large meeting and at the PCC (Program for Cooperative Cataloging) meeting on Sunday night. The aim of the summit was to bring together 70 people from all segments of the serials industry and the library community to determine the future need for bibliographic and holdings data in the management of electronic serials and to give CONSER advice on how to provide bibliographic records for electronic journals as well as on the full range of metadata for electronic journals. 

   Among the points that came up at the summit is that access is to the article level rather than journal level.  This led the summit participants to question whether you need a journal or would just a link from the article be sufficient.  Publishers thought that journals had a role for validation of content, browsability for current awareness, etc.  The role of OPACs (Online Public Access Catalogs) in finding journals is more ambiguous, however.  Someone said that about 80% of journal access came from outside of OPACs.  As for their view of print, publishers said there will continue to be print but it is more likely to be print on demand.  How important will volume and issue numbers be in this environment?  If journal titles remain important, how important is it to describe these titles?  More and more of us are looking to others to provide the data but they are looking at us to provide the CONSER records.  There is interest in subject access to journal articles but there are no standards for depth of subjects or on what terminology to use. One hoped-for outcome of the summit was cross-sector contribution to data creation and reuse.  Further exploration on these lines is needed.

    

   Another panel focused on standards. Lots of work is being done on syntax rather than content.  An important question is “Who owns the data?”  More and more data is being put into proprietary systems that don’t talk to other systems. Can you get the data out and re-use it?

    

   Karen Calhoun (Cornell) asked about the impact of the summit on CONSER.  Bob Wolven said that there had been about 15 recommendations.  Some were things PCC could deal with, others were concrete but not easy; others were more difficult.   Some of the things CONSER was looking to do are determining what data is useful in records and where is it being created?  They hope to get a consultant to look at all the data available through the summit.

    

   At this point Les Hawkins, the CONSER Coordinator, was invited to join the table.  He said that CONSER is looking to transform itself from a producer of records to a facilitator of uses of records.  One summit recommendation had been to create an advisory board for CONSER (composed of producers, vendors, etc.)  Another was to have CONSER provide better coverage of electronic databases. 

    

   Judi Nadler (Chicago) said it had been a very good conference. She saw CONSER as a program facilitating discussion of use of records. She wanted CONSER as a program to continue to maintain that dialogue.   Jeffrey Horrell (Harvard) asked about CONSER’s relationship to emerging institutional repositories.  Les Hawkins said that had been discussed. One question raised was whether these repositories were isolated islands, it was not yet clear how their data could be harvested.  It wasn’t yet settled who would do archiving of e-journals.  In the long run metadata for those journals would still be important.   Bob Wolven said there was a crying need for someone to take a leadership role in bringing the stakeholders together.  Karen Calhoun suggested using LC's approach to the action plan as a model for moving the action items from the CONSER summit forward.    Beacher Wiggins (LC) said the PCC steering committee had discussed how to carry things forward; they thought it would be desirable to hire a consultant to determine a time line and actions that should be taken. 

    

   Arno Kastner (NYU) asked Bob Wolven to speak more on the question of the content of records.  Wolven said there had been a feeling that not all existing data elements were needed. Judi Nadler asked about open URLs; Bob said that the ability to incorporate open URLs in CONSER records had been recommended.  Les Hawkins said that was proving rather complex for various reasons.  It might be possible to have publishers assign title level DOI (Digital Object Identifier), if an appropriate ISSN can be part of that assignment, and use DOI as part of a locally resolvable URL in CONSER records.  

    

   Karen Calhoun noted that our libraries are all preparing to load data in ERM (Electronic Resources Management) modules and this may represent a new opportunity for sharing data and reducing duplicative effort, provided we can agree on some standards or best practices.   She was interested in more than bibliographic descriptions and holdings information, such as sharable licensing information, information about aggregator packages, and so on.  Lee Leighton (Berkeley) said the University of California was looking at some ERM approach that would apply to all the University of California libraries.

    

   Bob Wolven (Columbia) said one interesting idea that came up was for a shared database that would contain core license terms from different publishers.

    

   Karen Calhoun suggested that those using ERMs might share their plans in at the January 2005 Big Heads meeting.  Jeffrey Horrell wondered if vendors might also be asked to contribute.    Someone suggested that we might incorporate vendors input in our reports.   Bob Wolven said he would be interested in knowing what topics people were discussing with vendors as this would give us an idea of what we could get out of the databases later.  Joyce Ogburn (University of Washington) said Tim Jewell would be interested in talking to us about what vendors are doing.   Catherine Tierney (Stanford) asked to what extent the DLF (Digital Library Federation) is doing things. Lee Leighton said they are doing a lot.   Bob Wolven noted that Indiana University is attempting to fill in CONSER records for things that aren’t there.  He noted that lots of libraries are canceling print subscriptions and wondered how they were keeping up with changes in ejournal titles, etc.  John Riemer (UCLA) said UCLA was asking vendors to give it frequent lists of coverage that they could compare for changes. Other Big Heads agreed they wanted something similar.

    

   Arno Kastner asked how people were providing subject access. One approach was to map the LC class number.  Bob Wolven said Columbia had provided its data to Cornell. Karen Calhoun said Cornell was using the Columbia data for developing an online menuing system for subject browsing.  John Riemer asked if they were finding this use of LC classes helpful for accreditation but there was no direct answer.  Bob Wolven said it might be useful to see what people are doing and the use they are making of such mappings.

    

   It was suggested that Electronic Resource Management systems be placed on the agenda for the January 2005 Big Heads meeting.  John Riemer asked if ISSN data might be verified and assigned.  Les Hawkins said that CONSER is working with the ISSN Network on ways to provide ISSN data for records that lack it.

    

   Richard Jasper (Wayne State University) asked, from the audience, what people were doing re-assigning staff to deal with electronic resources matters.  Up to now this work has been done by high level staff but he now wondered if it could be pushed down.  Lisa German (Illinois) said yes.  Joyce Ogburn (Washington) said she doesn’t want staff to deal with it, she wants automation to do it.

    

   Someone from the audience said that NISO (National Information Standards Organization in the U.S.) and EDItEUR (an international group coordinating development of the standards infrastructure for electronic commerce in the book and serials industries) are developing an ONIX (Online Information eXchange) to facilitate distribution of this information.

    

Break: 10:40 - 11:05 a.m.

5. Vendor records for set analytics:  increasing their availability, improving their quality, agreeing on a reasonable price.  Is the ECCO set a model?   What about Source OECD, Jane's, AltHealthWatch, Alexander Street Press sets, etc.

   Housekeeping issues: Arno Kastner announced that Becky Lyons, NLM Acting Chief of Technical Services and Caroline Early, Acting Head of Technical Services for NAL, were representing the National Library of Medicine and the National Agricultural Library respectively at this meeting.

    

   Joan Swanekamp (Yale) noted that vendor-supplied records have been a fixture of bibliographic control for a long time.  Key points for such records are availability, quality, and price.  We can’t buy records that don’t exist and we won’t buy records that do not help users gain access to the materials they describe.  We have to have records that sustain the relationship between buyer and seller; if the price is too low no-one will make records but if the price is too high, we will not buy the records.

    

   Yale’s experience with digital product suppliers has been positive.   Overall, vendors have been responsive to library needs.  As an example, she described their experience with the Eighteenth Century Collection Online (ECCO), an online collection which parallels a microform collection, the Eighteenth Century Collection. The records lack much subject analysis.  Yale wanted to load records, and add subject analysis as a later phase.  They saw a similar pattern with other vendors who provided records: the records had catalog data they did not need, non-MARC21 tags, some had notes relating to the microform version.  Yale had the originals but use of records for the microform versions would make it look as if Yale had the microform duplicates instead.  They needed records with 007 and 006 fields, coded correctly.  They needed much interactive work with vendors to get what they needed.  Yale worked with Gary Strawn (Northwestern) and Margarreta Yarborough (University of NC at Chapel Hill).

    

   Trying to direct vendors towards the PCC website (http://lcweb.loc.gov/catdir/pcc/aggtg2final.html ) where several task forces of the PCC Committee on Automation have provided a model for record content and format seems a better solution than individual negotiations with vendors.

   At Yale more than half an FTE of a high level professional for the past year has been spent working with vendors (ECCO, Evans, and other smaller vendors) on specifications.  Some of the smaller ones didn’t have records in MARC.    Judi Nadler said that at Chicago this work is also done in-house; the work is shared between the head of Database Management in the Cataloging Department and someone in the Integrated Library Systems Department.  She agreed that a centralized approach would be preferable.  Bob Wolven asked what PCC could do to improve this process.  Joan Swanekamp said they just needed to make the PCC Standing Committee on Automation’s e-aggregation model better known. Sally Rogers (OSU) said that OhioLink has been using the PCC site to develop record standards that vendors will be asked to follow at the point of contract negotiation for purchase of their electronic products.  She will find out what additions or changes to the PCC site information were considered necessary.    Irene Zimmerman (Wisconsin) asked if we knew if the vendors were sharing the changes they make for us with other subsequent clients.  Joan said she thought the large vendors were.  

    

   Arno Kastner said there were opposite instances in which, in response to a request from a single library, a vendor made across the board changes that are not suitable for everyone; joint discussion with vendors is needed.  Bob Wolven said this was an example of weakness in the PCC site: it had standards of what should be included in records but didn’t say what should be deleted, e.g., microform notes from records that were being changed to describe electronic resources.  He commented that it is also important to tell vendors which of the things you are requesting that are for your specific needs and that probably would not be useful for other libraries. 

    

   Karen Calhoun (Cornell) noted that libraries are also suppliers: they are sometimes asked to provide records for their digital collections. Cornell makes selected record sets available for download from its Web site, but as is, since the library lacks the resources to customize them for use at other libraries.  Michigan is taking the same approach.   Margaretta Yarborough (NC) said they are providing record sets to OCLC but users can modify the records.  Bob Wolven said that Columbia provides subscriptions to records for CIAO (Columbia International Affairs Online) and other collections.   Mechael Charbonneau said that Indiana has a fulltime programmer within Technical Services who works with vendor-supplied records. In a joint project with the University of Michigan, microform records for Wright's American Fiction, 1774-1990, were used to programmatically create new records describing the digitized version.  Arno Kastner asked how people were tracking what record sets for electronic resources are available; he said that at NYU they are compiling a spread sheet to maintain this information.   Joyce Ogburn (U. of Washington) recommended that those who create bibliographic records for their own material put these records in some central source, such as OCLC, where others can use them. Lisa German (U. of Illinois) noted that different ILS vendors require different information.  She urged that everyone who had scripts they had prepared for use in their own systems should publicize their availability.    Bob Wolven said that Columbia has started doing some analytics of sets.

    

   Catherine Tierney (Stanford) commented that once you get these records into your catalogs you then can't export them to share with others.    Nancy Gibbs (Duke) said that such a prohibition was in a separate license agreement that Duke had signed but that they crossed it out.  Catherine Tierney said ECCO’s license says you can’t export the records.

    

   Arno Kastner (NYU) noted that there used to be dialogue with vendors that tried to convince them that the availability of records in OCLC sets was to their advantage, that it would encourage more libraries to purchase the sets.   Joan Swanekamp said the cost of many of these sets is quite prohibitive. Jeffrey Horrell (Harvard) pointed out that if you own much of the original material you can negotiate.

    

   Judi Nadler (Chicago) emphasized that we need to centralize what is available, perhaps at PCC.

   Phelix Hanible (Michigan) asked if we had information on user behavior: do they come to the catalogs to find this material or to web sites?  Judi Nadler said she thought the catalog.  Catherine Tierney asked if PCC might look at the issue of exporting records, advising publishers that it would be to their advantage to allow libraries to do so.   Beth Picknally Camden (Virginia) said if we had data on use of URLs we might have more clout. 

    

6. PCC records--have they made a difference in our processing; have they raised the overall quality of access?

   Arno Kastner (NYU) said that a survey that Big Heads had done 3 years ago showed that libraries were making quite different use of PCC records.  He wondered how high volume libraries are using them today. 

    

   John Riemer  (UCLA)  said that they are very valuable, providing access to high quality records, akin to those produced by the Library of Congress.   Irene Zimmerman (Wisconsin) confirmed this, saying that Wisconsin had found that LC records needed 1.3 edits per record, while PCC records needed 1.8.  This compared to the 4.3 edits needed for general contributed cataloging.

    

   Lee Leighton (Berkeley) agreed that Berkeley found PCC records very useful; they just had to add LC call numbers to some of them. Irene Zimmerman said Wisconsin chiefly had to add publication dates in the 008 and 260 fields..

    

   Arno Kastner asked if staff had been asked to look at PCC records and pass them through more quickly.  Irene Zimmerman said Wisconsin was now grouping records with call numbers in the 050 field versus those with call numbers in an 090 field.

    

   Judi Nadler said that the University of Chicago is taking a more extreme approach.  Both DLC and PCC records are completed in the Acquisitions Department.   This approach has proved popular with the library administration and public services staff but not with Catalog Department staff.    The major value of PCC records is to be able to route them that way.   She thought more libraries should contribute them and more should use them. 

    

   Cynthia Clark said that NYPL is uses PCC records as they do those from LC, though they have to do some editing of both because of their local classification system.  Katharine Farrell said Princeton also treats DLC and PCC records in the same way, using the lowest level of staff.  The records are not edited.  She noted, however, that the quality of authority control on PCC records is lower than that of DLC and that some headings may not be under authority control.   Sally Rogers said that Ohio State had found similar problems: e.g., series not established.   Bob Wolven said a problem was that we can’t agree on what is acceptable.  Judi Nadler agreed but said it might be easier to agree on what is required and how to achieve the right balance between not inhibiting libraries that want to participate and maintaining quality.

    

   Karen Calhoun said that at Cornell, as at Chicago, both PCC and DLC records are treated in  the Acquisitions Department. Patrons and public services staff are happy to have materials coming out faster and that too is an aspect of quality of access, as is the ability to reduce cost so money can be re-allocated to other areas of the library such as original cataloging.   The possibility of original catalogers becoming trainers is also a positive thing. 

    

   Beacher Wiggins (LC) commented that catalogers are suspicious of some PCC records and that is an attitude that needs to be treated.   As Catherine Tierney had pointed out, we have so much else to deal with we can’t afford to re-inspect PCC records. 

    

   Bob Wolven said that if no more PCC records appeared it would make little difference to Columbia’s workflow.  Beth Picknally Camden said that Virginia treats PCC records like member copy; they felt they couldn’t introduce another workflow but the presence of  subject headings and class numbers in PCC records makes them go through more quickly.

    

   Catherine Tierney (Stanford) suggested that the Big Heads encourage their staff to create PCC records because that enables us all to do our work better by getting more records through quickly.

    

   John Riemer (UCLA) asked if there is an attempt to recruit more BIBCO libraries.  Both Bob Wolven and Beacher Wiggins said that a campaign to add more wouldn’t increase contribution very much.

    

   Judi Nadler (Chicago) noted that she now has copy catalogers who can do lots of other things besides copy cataloging.

    

7. Searching across digital collections: federated searching? OAI repository?

   Karen Calhoun said that Cornell got into the digital collection building business about 10 years ago and now has lots of digital collections. As they built these collections they took a project rather than a programmatic approach.  Because these collections are scattered over their website people have difficulty in discovering what collections Cornell has and what is in them. Now have the objective of creating an integrated technological and methodological framework for making the library's many digital collections more easily and conveniently discoverable.

   The first phase (dealing with licensed collections) was completed last summer; it introduced federated searching and reference linking.  They next turned to collections digitized by the library and are investigating 3 issues:

   1) The number of delivery platforms they have.  They have 6-7 of them and wonder if they could go to fewer and make better use of those platforms.

   (2)  Identification of a set of best practices and standards and assumptions underlying collection development policies.

   (3) An appropriate model or models of collection building. They are looking at an OAI repository model (many collections are not harvestable) and also a federated search model. Questions of interoperability have been raised.

    

   Joyce Ogburn (Washington) said she is torn on these issues. Scholars work with whatever platforms work best for them and we have to adapt.  But there are too many platforms we create in the library ourselves.  Rationalisation of collection development is needed.  Washington primarily uses ContentDM.   Joan Swanekamp said that Yale’s situation is similar to Cornell’s with lots of entrepreneurial projects which then come to the library to be preserved.  She thought that Cornell was asking great questions.   Jeffrey Horrell (Harvard) said it isn’t just in-house collections that matter, but also the ones we get from vendors.  We need to be able to provide users with access to content.

    

   Cynthia Clark said that NYPL is dealing with similar issues:  They have digitized audio and video files and now have hundreds of thousands of images that have been scanned; they now have to integrate these materials into their catalog.

    

   Beth Picknally Camden said that the University of Virginia has been funded by the Andrew W. Mellon Foundation to enable the Library in collaboration with the Cornell Computer Science Department to build a digital object repository system based on Fedora (Flexible Extensible Digital Object Repository Architecture). The aim of the project is to provide one place where all digital objects can be found and searched. (See http://www.fedora.info/history.shtml)   They are currently working on a discovery index for the repository.  Objects will retain their native richer metadata but will also have a common index. Virginia expects to go live with it at the end of summer 2004.

    

   Bob  Wolven said it was less important for scholars to find one interface for Columbia’s digital collections than to be able to locate similar materials in all its many other collections from many different libraries.    Karen Calhoun (Cornell) mentioned TEL (The European Library), a cooperative of European national and national-like libraries that will offer integrated access to the collections of the libraries. (See http://www.europeanlibrary.org )

    

   Judi Nadler said that the University of Chicago is still at the project level stage. She asked who in your libraries is part of the dialogue.  The consensus answer was:  Everybody. To the question of who coordinates the everybody, the answer was that it often was one of the Associate/Assistant Librarians, working closely with Information Technology people.

    

   Joan Swanekamp said that Yale has lots of digital information and is concerned about long-term preservation issues.

    

8. Acknowledgments.

   Arno Kastner in his role as Chair acknowledged the 15 years of service that Judith Hopkins (University at Buffalo) has provided to Big Heads as their recorder.  On her retirement they presented her with a pin from Appalachian Spring and a printed tribute that read as follows: Judith Hopkins, In honor of your fifteen years of service as recorder for the ALCTS Big Heads Discussion Group, the Directors of Technical Services for Large Research Libraries salute you.  You have been our faithful scribe and aide-de-camp. So doing you have ensured that our words live on through your words.  You may have been an honorary Big Head, but, to us, one and all, you will always be remembered as a true Big Heart.  We express our deepest appreciation for your many years of honorable and generous service and wish you in retirement an elegant sufficiency.

   Arno Kastner acknowledged Rick Schulz (Princeton) for the tribute, Robert Milevski (Princeton) for creating the tribute on a handpress in the Princeton University Library, and Scott Husby (Princeton) for creating the presentation folder.

   Judith expressed her gratitude and described how her work as Big Heads recorder had grown out of a personal defect, her inability to remember what she hears.  To compensate for that defect she had started taking detailed notes as a member of the Big Heads audience and then had shared them with other attendees.  Big Heads then asked if they could use the notes as their Minutes.

 

The meeting was adjourned at 12:35 p.m.