American Library Association (ALA) Annual Conference

Citation

(2002), "American Library Association (ALA) Annual Conference", Library Hi Tech News, Vol. 19 No. 8. https://doi.org/10.1108/lhtn.2002.23919hac.008

American Library Association (ALA) Annual Conference

with contributions by nine authors

Introduction

The following is a sampling of the highlights of the 121st American Library Association (ALA) Conference held in Atlanta, Georgia from 14-19 June, 2002. This conference is the world's largest and most comprehensive library conference and exhibition. The ALA Conference includes more than 2,000 meetings, discussion groups and programs on topics affecting libraries.

This year with have three reports from the Reference and Users Services Association (RUSA). Pam Werre provided an in-depth report on the RUSA/MARS preconference on digital reference. Susan Thompson prepared an extensive report from the Machine-Assisted Reference Association's (RUSA) Local Systems and Services Committee on usability testing. Claudia Timmann wrote about the thought-provoking RUSA President's Program on reference interviews in cyberspace.

Our next group of conference reports comes from the Library and Information Technology Association (LITA) presentations and programs. Betty Landesman reported on the LITA preconference about a new standard the XML – eXtensible Markup Language and how it can be used in library applications. Stacy Voeller prepared a report on the LITA President's Program about electronic preservation in cultural organizations, specifically the National Cultural Heritage Network (NINCH). Susan Thompson wrote for us again on the program presented by a new LITA Interest Group, the Heads of Library Technology, on the challenges of recruiting and retaining library technology staff. Megan Kinney reviewed the latest program from the Open Source Interest Group on major open sources Web finding tools. Our ALA reports conclude with a report by Jonathan Rothman on the informative program from the Technical Issues of Digital Data Interest Group on preserving access to digital collections.

For the 2003 Annual Conference, the ALA will be holding a joint conference with the Canadian Library Association. The 2003 ALA/CLA Annual Conference will be held in Toronto, Ontario from 19-25 June, 2003.

Pam Werre

Are You Ready for Digital Reference?

The RUSA/MARS preconference held Friday, 14 June, 2002 titled, "Digital Reference @ Your Library" was a jam-packed day of practical advice, philosophical discussions, and words of wisdom from those experienced in the area of virtual reference. The day was divided up into sections with presentations grouped into appropriate categories. The session began with keynote addresses from R. David Lankes, Sara Weissman, and Bernie Sloan. Lankes opened with a list of milestones to indicate that virtual reference (VR) has arrived as a library service and what he expects will happen in VR in the coming year. He stressed that librarians should view digital reference as an opportunity to re-introduce and re-energize reference service. Weissman suggested calculating service rates for traditional, email, and virtual reference to determine what services are feasible and important to continue. Her formula for calculating the service rate was to divide the number of questions in a given time period by the service population. For example, the service rate could be calculated by dividing the number of VR questions in a month by the number of students and faculty at an institution. This service rate could then be compared to the service rate for email reference and traditional desk reference. Sloan raised discussion questions for attendees to ponder such as "where does digital reference fit in with other reference services" and "is your VR service designed for librarians or for the end-user" and "how do we set goals and benchmarks for accessing virtual reference"

Sam Stormont gave one of the most useful sessions for those librarians just starting a virtual reference service. Stormont presented five models of VR: basic, homegrown, advanced, collaborative, and corporate contact center with definitions for each model. Stormont outlined a ten-step plan for implementing a live virtual reference service:

1. 1.

   articulate a vision;

2. 2.

   form a committee;

3. 3.

   gather information;

4. 4.

   investigate funding;

5. 5.

   decide on staffing;

6. 6.

   select software;

7. 7.

   draft a policy;

8. 8.

   develop a marketing strategy;

9. 9.

   implement LVR; and

10. 10.

    evaluate LVR.

The preconference included sessions on both national and local virtual reference projects to give the audience both a global and a local perspective. The next two preconference presentations highlighted national projects. Nicholas Savard presented a session on digital reference at the National Library of Canada (http://www.nlc-bnc.ca/index-e.html) and Diane Kresh presented a session on QuestionPoint (presentation online at http://www.loc.gov/rr/digiref/) the collaborative reference service hosted by the Library of Congress and OCLC.

Diana Sachs and Susan Herzog from the Public Library of Charlotte and Mecklenburg County ()http://www.plcmc.lib.nc.us/) presented the local project perspective. The librarians at this public library have partnered with the 24/7 network to offer digital reference at their public library. By contributing ten hours of staffing a week to the 24/7 service (http://www.247ref.org/), their community receives access to a network of reference librarians who are available 24 hours a day, seven days a week to answer questions.

Every preconference needs a radical librarian to shake up the crowd and to offer an alternative viewpoint. Rob Casson from Miami University was both engaging and cutting edge. Casson described his project of creating and writing the code for a chat reference service at Miami University called RAKIM. This is an open source code that Casson is updating and improving and is available on the Web at: http://styro.lib.muohio.edu/rakim/.

R. David Lankes made a repeat appearance to address the group on quality and performance in virtual reference service. Lankes has been involved in a research project to assess quality in digital reference (http://quartz.syr.edu/quality/). To assess VR librarians should consider both performance and quality standards. Lankes defined performance standards as descriptive statistics, log analysis, cost analysis, and user satisfaction measures. Quality standards were outlined by Lankes to include accuracy, user satisfaction, awareness, and number of repeat users.

Linda Klimczyk, Penn State, reminded the audience to design accessible electronic and digital reference sites in her presentation, "Access and identity in e-reference." Libraries should take care to design sites that reduce barriers so the service is usable by anyone with a disability or communication disorder.

The virtual reference interview was the hotly debated issue of the day. Matthew Marsteller and Danianne Mizzy from Carnegie Mellon discussed the perception that digital reference is suited only to answering short questions and that patrons do not have the patience for real reference interviews in the chat environment. They conducted a study of VR transcripts and analyzed the questions and responses for a one-year period. They concluded that patrons are asking significant numbers of in-depth reference questions and that patrons have an overwhelmingly positive response to librarians asking questions to clarify their informational requests. They further concluded that the reference interview has a place in VR, especially for more in-depth research type questions.

Lisa Horowitz and Sarah Wenzel from the MIT libraries focused on the practicalities of staffing and implementing a VR service. They approached the process from a scientific standpoint and encouraged librarians to apply a scientific methodology when implementing a new service. To accomplish this, Horowitz and Wenzel encouraged librarians to articulate the problems, evaluate priorities, create a working hypothesis, and develop an experimental model for the service. By working through the process scientifically librarians give themselves the freedom to experiment with digital reference, re-evaluate staffing and priorities, and allow themselves room for failure if the experiment is not successful.

The preconference concluded with presentations on four currently operating services. M. Kathleen Kern and David Ward described implementation of digital reference at the University of Illinois at Urbana-Champaign utilizing three reference desks and graduate assistants in addition to librarians. Janie Silveria, California State University, Monterey Bay, described their collaborative public and academic library project called QandACafe in northern California. Key points were the length of time to answer questions (10-25 minutes per question), the difference in public library and academic library type questions, and suggestions for best practices in answering questions. Glenda Schaake and Eleanor Sathan described how their small public library in Andover, Massachusetts, offered virtual reference service as a part of the 24/7 collaborative network. Maureen Morris and Virginia Cole from Cornell explored the feasibility of a coast-to-coast collaborative VR project between the University of Washington and Cornell University. One of the key outcomes of their project was the discovery that it was practical and feasible to collaborate on a digital reference project and that librarians at other institutions were able to answer questions from patrons at their home and partner institutions.

Poster sessions and digital reference product demonstrations rounded out the offerings of this preconference. ExLibris, IBSI LiveAssistance, OCLC, 24/7 Reference, VRD had booths with knowledgeable representatives and pricing information. Seven poster sessions with a variety of topics and materials were available in the room to visit during breaks. Poster session abstracts are online at: http://www.ala.org/rusa/mars/preconfbios.html#posters.

This preconference was carefully constructed with speakers and presenters well versed in the virtual reference arena. The sessions provided practical and relevant information for librarians in all stages of implementing digital reference; from those still considering the service to those who are already offering VR in their libraries. The preconference agendas are available on the RUSA Web site at: http://www.ala.org/rusa/mars/preconf.html

Pam Werre(werrepa@mnstate.edu) is a Public Services Librarian at Minnesota State University Moorhead, Moorhead, Minnesota, USA.

Susan Thompson

Usability Testing for Reference Librarians

This is a report on the Local Systems and Services Committee's Machine-Assisted Reference/RUSA program entitled, "Usability testing for reference librarians: doing it cheap, doing it right, doing it yourself." Reference librarians are often involved in the design of Web pages and database interfaces. User testing of interface design is vital to this process. The speakers discussed the principles and techniques of usability testing based on their experience and expertise with an emphasis on how testing can be done locally and inexpensively.

Teal Anderson a usability specialist at the Digital Knowledge Center Sheridan Libraries of John Hopkins University led off the program with a basic overview of usability testing. Her definition of usability included measurable characteristics such as ease of learning and use, effectiveness, and satisfaction. In practice, usability focuses on determining what the users' needs are by observing users interacting with the site. Usability tests are important for the insight they provide into problems with the Web site interface and possible solutions. The importance of usability testing was summarized by a quote from Susan Dray, "if the user can't use it, it doesn't work!"

Testing should take place before you design your site, while you are redesigning the site, and after you design your site – in other words, throughout site development. A variety of evaluation methods can be used. Heuristic evaluation and cognitive walkthroughs are not strictly usability tests because they do not directly involve the user. They do however provide formal ways for the expert Web developer to examine the site from the user's point of view. Sorting and naming tests are done with users. They help determine how to categorize and name the information content in the site. Focus groups, interviews, and surveys directly ask the users' opinions and are a good way to find out what users want, including new features.

Observation of the user interacting with the site is the classic form of usability testing. Teal described two types of observation – naturalistic and think-aloud, scenario based. Naturalistic observation takes place in the user's own setting and passively observes how successful the user is in carrying out their normal tasks. Think-aloud, scenario-based tests are more controlled observations where the test administrator sets up the testing situation such that the user carries out predefined tasks that will examine how successful they are in using specific features of the site. Ideally, both forms of observation should be used.

Teal described how to conduct usability tests using her P⁴ model (see Figure 1 ).

1. 1.

   People are the most important part of the test. Recruiting a representative sample of users is ideal but not always realistic. Useful data can be obtained even with a self-selected group such as volunteers responding to email or posted solicitations. It is important to treat test users with respect including how we refer to them – as "participants" rather than "subjects" sets the appropriate tone. The other important people in the process are the test administrators. She recommends a facilitator, who interacts directly with the user, and an observer, who takes notes on all the user's actions and comments.

2. 2.

   Paraphernalia involved in the testing process include the computer, equipment to record the session, and incentives. Recording can be as informal as pen and paper but audio/video or other equipment can be used. However, one should bear in mind the time spent in transcribing from other media. In fact, Teal recommends taking observation notes directly on a laptop to avoid even the need to transcribe paper notes. Incentives help bring in volunteers; however, they should be geared to the type of participants. The coffee and pizza coupons that bring in students may not work as well to attract faculty to the test. She found food was good for focus groups and a $10 gift certificate at the campus store worked well for individual participants.

3. 3.

   The testing place depends on whether the observation is to be naturalistic or controlled. Controlled observation, particularly with the think-aloud protocol, works best in a quiet room or office.

4. 4.

   The final part of the process is the paperwork and it is also very important. At the beginning of the process, many institutions require some sort of informed consent and internal review board approval. Pre-tests, usually to gather demographic information, and post-tests, such as a Likert scale to determine level of satisfaction, are typically used and are helpful in interpreting results of the usability test. The test administrators should use a test script, from welcome to thank you, to ensure consistency across participants. A test log provides a way for the observer to keep track of everything going on and might include checklists on whether users perform common actions as well as providing space for notes. After all the data has been gathered, a summary/analysis is undertaken to look for trends, particularly convergences, and for places where people say one thing but do another. Highlights of the key findings and recommendations of specific actions to be taken are presented in the final report.

Figure 1 the P⁴ model

Mary P. (Molly) Freier a research and instruction librarian at the Gould Library of Carleton College discussed the Carleton College library Web site redesign and provided insights into usability testing from a reference librarian viewpoint. Her insights as a novice usability tester pointed out assumptions someone experienced with usability may have that others in the library or on the committees involved in the process might not share. For instance, she had gone into the testing expecting to see consistent, clearly delineated results that would make selection of the best interface choice not only obvious but provide easily supportable evidence to back decisions in the political arena. Instead she found the results of usability tests were much fuzzier and more subject to interpretation than expected.

For instance, users clearly indicated a preference during testing for a new design using only three entry points as compared to the original site, which had 50 links. Confusing the results, however, was the fact that users completed the tasks using the original site much more quickly. In trying to understand why users would prefer a site, which, on average, took longer to use, she analyzed the time statistics for each of the three interfaces in several ways. No clear trends were obvious until she compared the maximum amount of time it took to complete the task. The three-entry point design was dramatically faster, especially given the fact that most of the user's were familiar with the existing 50-link site. The reason lay in the fact that the user never got "stuck". It was easy to figure out where to go with the three entry design and false starts were minimal since the user had only two other choices to try.

On the other hand, she found that it was important to keep features of an existing site that are working. The one feature of the 50-link site that users did appreciate was the ability to get to their destination in one click with no intermediary navigation pages. The final site design incorporated this one click capability by providing each of the three entry points with rollovers containing extensive lists of choices on popup menus.

Another interesting finding that Freier talked about was the use of pictures – a subject not commonly covered by library usability tests. The Carleton team found that students preferred photos on the site, something faculty was much less enthusiastic about. Perhaps because they were so interested in photos the students were also very sensitive about the content of the pictures, particularly if they looked even slightly dated. They did not like "posed" pictures or pictures of things like sleeping students or even a girl with unfashionably bright lipstick. As a result, photos were used throughout the final site but "doctored" where necessary. For instance, they kept the picture of the girl but used Photoshop to tone down the lipstick to a more acceptable shade in the final site.

Martin (Marty) Courtois a reference/instruction librarian at the Gelman Library of George Washington University described the Web usability testing at the library. During the Gelman Library Web redesign project, the team had several goals they wanted to achieve with usability testing:

• •

  observe users interacting with the site, preferably in a natural setting;

• •

  target freshmen students;

• •

  look at how to both make the site easier to use and more instructive in the use of library resources.

The team planned to conduct 15 tests. They kept the tests short, 30-45 minutes, by limiting to 12 questions. The guidelines they used in formulating test questions was that they were scenario-based, defined a specific information need, were easy to understand, and easy to score. The questions focused on determining if users could find basic information and services, such as the OPAC, databases, and library hours and if the users could find information on how to use library resources, such as how to find periodicals or how to use reference books.

Before creating their own questions, the team looked at a variety of questions from tests conducted by other libraries. They noticed several characteristics in these tests that they did not like. Many questions seemed to test the user's ability to use the OPAC or conduct some other form of library research. The Gelmen team was more interested in how the user navigated through the site than how they used the resource once they found it. They also noticed that some of the questions in other studies were open-ended which made them both harder to score and unclear as to when the task was accomplished. Based on this analysis of other test questions, the Gelmen team developed several guidelines for their own questions:

• •

  The question should not require the user to go into the catalog, database or other information resource. The task was completed when they found the page to start their search.

• •

  They decided not to ask users to think out loud. They felt that it created an unnatural condition, particularly in that users were more likely to persevere long after they would normally have given up. They also felt thinking out loud would be uncomfortable and interfere with accurate timing tasks. However, they did decide to do focus groups afterward to try to elicit student opinions.

Specific instruction were given to the students at the beginning of the test:

1. 1.

   start from the homepage;

2. 2.

   find the page where the resource is located (but do not go into the resource);

3. 3.

   use any method desired as long as you stay within the library site;

4. 4.

   tell the facilitator when you think you have found the page (demystifies where the user thinks they are in the process);

5. 5.

   a three minute limit on each question plus user was allowed to quit at any time;

6. 6.

   the facilitator emphasized that it was the site being tested not the user.

Conducting the user tests was made much easier by the cooperation of a professor who gave extra credit to students in his Internet class who participated in the usability testing. The team gave the test participants a pre- and post-test. They found that they did not really use pre-test results but the post-test was very helpful. It basically helped them debrief participants with open-ended questions that elicited the user's opinions – likes and dislikes – on the site.

Results from testing showed that users generally were able to find basic information about the library and the key resources easily within a few clicks. On the other hand, it was very hard for them to find information on how to use the library and its resources. As a result the Web team created a pulldown menu with links to "how to" pages.

Their observations during testing showed that users rarely read the content of pages carefully or used the navigation bars (the back button seemed to be all they needed). The team concluded that homegrown usability testing and focus groups, even imperfectly done, revealed areas of the Web site that needed improvement. They found they could do the testing on a small budget. They felt it was valuable to test discrete user groups. Finally, usability testing does not obviate the need to study good Web design, it only confirms if it is successful.

For more information see the program Web site: http://www.ala.org/rusa/mars/usabilty.html

Susan M. Thompson (sthompsn@csusm.edu) is the Library Systems Coordinator at California State University San Marcos, San Marcos, California, USA.

Claudia Timmann

The Reference Interview

At the American Library Association's (ALA) annual conference, the Reference User Services Association (RUSA), a division of the ALA, hosted the annual President's Program on Monday, 17 June 2002. RUSA is responsible for stimulating and supporting every type of library in the delivery of reference/information services to all populations. This involves facilitating the development and conduct of direct service to library users, the development of programs and guidelines for services to meet the needs of the users and assisting libraries in reaching potential users.

The title of the program was "The Reference Interview: Connecting in Person and in Cyberspace." It was held at the Sheraton Atlanta Hotel and involved more than 120 academics and information professionals/librarians from around the world. The focus of the program was the face-to-face interaction of the reference interview versus the invisible online/virtual interaction.

There were two speakers and three reaction panelists. The first speaker was Catherine L. Ross, professor and dean of Information Media Studies at the University of Western Ontario in London. Ross gave a humorous analysis of the reference interview by sharing her students' experiences at the reference desk. She mentioned that librarians and information providers do not realize that "users don't even know what reference means, let alone indexing elements." Frequently librarians and information providers assume users know how to ask for the information they need. Overall, the librarian and the reference interview are the keys to the information seeking process for users. The solution to connecting users to information is not to reduce the human element, but to make it stronger and more effective.

Another engaging speaker was Jana Ronan, the Interactive Reference Coordinator for RefeXpress, a chat reference service at the University of Florida. Ronan provided suggestions of effective online reference interviewing. She stated that one of the biggest challenges with providing virtual reference (or real time reference) is learning how to communicate effectively with remote users and how to translate personal skills used at the physical reference desk into the virtual environment.

The first panelist, M. Kathleen Kern, Assistant Reference Librarian at the University of Illinois at Urbana-Champaign, resoundingly stated three main points in regards to virtual reference, librarians/information providers and users:

1. 1.

   speed of delivery is not the defining characteristic of virtual reference services;

2. 2.

   virtual reference services are in the physical library just as in-person reference services are located in the library; and

3. 3.

   virtual reference gives librarians and library providers opportunities to re-examine and realign service philosophy.

The second reaction panelist, Dr Marie L. Radford, acting dean at the School of Information and Library Science at the Pratt Institute in New York, provocatively gave recommendations in terms of the emergence of virtual reference and the needy user population. She pointed out that librarians and information providers have come to a critical turning point in reference service and that librarians must:

• •

  embrace the change and redesign services to support the needs of users;

• •

  collect statistics differently; and

• •

  consider chat services that offer co-browsing, escort assistance, instruction (tutorials), chauffer and modeling – as the librarian/information provider does a search the user can learn by watching.

The last reaction panelist, David Tyckoson, head of reference at the Henry Madden Library at California State University-Fresno, focused on the face-to-face interactions users have at the reference desk. He stated that the reference interview is librarianship and that users need people they can approach. He said that librarians are information counselors – ones who guide people to the path of information.

Also, he was passionate about the reference interview by pointing out that "reference is the heart of what a reference librarian is." He strongly believes that users focus on the human interaction, not the facts that were given during the reference interview. Technology and the way the world communicates will change, resource tools will change and vendor sources will change, but the reference interview process is the core thing that will make successful reference librarians and happy, satisfied users.

There were only two responses from the audience. A librarian stated that she agreed with all that was said. However, she did not agree that librarians have to conduct the reference interview, which involves standing at the physical desk or sitting in front of a computer engaged in virtual reference/real time reference.

She emphatically stated that paraprofessionals could also be hired to do the same thing – to answer questions and conduct the reference interview. Another librarian mentioned that he never experienced chat or virtual reference, but felt that there would be a loss in the face-to-face interaction. He asked everyone "Can you get the same rush and the same connection with the user – the rush one gets when one knows one has helped someone – can this be done online via virtual reference?"

The panel and speakers did not have a response to the first statement, but did respond to the second by simply saying yes.

Claudia Timmann(timmannc@u.library.arizona.edu) is an Undergraduate Services Librarian at the University of Arizona, Tucson, Arizona, USA.

Jennifer Ward

Testing for Accessibility

The following is a report from the Accessibility of Online Library Resources for Users with Disabilities at the American Library Association (ALA) Annual Conference entitled "Testing for accessibility: throw your mouse on the floor and turn those images off!" Axel Schmetzke, Reference Librarian and Coordinator of Instruction and the Instructional Materials Center at the University of Wisconsin-Stevens Point, began the session by introducing the panelists and providing an overview of the topic. He noted that much of the information from this session will be covered in two upcoming Library Hi Tech special-theme issues. Mr Schmetzke also discussed his recent research on the accessibility of numerous library and library school Web pages noting the Priority 1 errors generated by Bobby (http://www.cast.org/bobby/), an automated accessibility checking tool. Approximately 80 per cent of the accessibility problems were images without "alt" text – something easily corrected, easily overlooked, and a large barrier to accessibility.

Judith Dixon, from the National Library Service for the Blind and Physically Handicapped, then demonstrated the (in)accessibility of several library websites and library catalogs using the JAWS screen reading software. She provided a quick and dirty way for the sighted to test website accessibility: "throw your mouse on floor, turn images off in the browser, and see how well you can maneuver". People with disabilities use dozens of keyboard shortcuts within the Internet Explorer browser or the screen reader to maneuver a site. Ms Dixon also suggested that accessibility assessment be done in teams of one sighted and one non-sighted person. The reason for this is expertise – one can clearly see what is on the screen and the other can easily hear what the screen reader thinks should be there. As she sagely noted: you don't know what you can't see. In addition to adding "alt" tags for images as a way to improve accessibility, she also recommended designers use meaningful web page titles. Ms Dixon did caution against relying wholly on Bobby to ensure accessibility, noting that it is better at finding instances of inaccessibility.

Mary Beth Chambers, Catalog/Archives Librarian from the University of Colorado at Colorado Springs, evaluated accessibility and usability of web based library databases. She tested Periodical Abstracts from OCLC Firstsearch and Expanded Academic ASAP from Gale InfoTrac using JAWS for Windows 3.7 and Window Eyes 4.1. Both are popular assistive technologies and readily available in her library. In this study, sighted librarians used screen readers to find problematic websites and then enlisted the help of non-sighted users to run a second test. Researchers examined the overall navigation of the site as well as the search and help screens. Both sites do meet most Section 508 (of the Rehabilitation Act Amendment of 1998) accessibility guidelines, although Ms Chambers noted that the majority of problems they ran into were usability rather than accessibility. She stated that usability of a product is an important factor in measuring its accessibility.

Sue Byerley, Electronic Reference Librarian from the University of Colorado at Colorado Springs, summarized a total of five studies on accessibility of online library resources for users with disabilities. In total, the studies looked at 15 different search interfaces provided by a variety of vendors and researchers used a variety of adaptive technologies to evaluate the resources. All studies had the same basic goals of examining the accessibility and usability of site navigability, basic and advanced search screens, help screens, and task buttons. Most resources were accessible by one means or another, but five earned the lowest marks of "accessibility significantly reduced" – EBSCOhost, Electric Library, Proquest, PsychInfo from Silver Platter, and PubMed. Kudos ("very accessible") went to Lexis-Nexis, Medline from Ovid, and the Social Sciences Citation Index from Web of Science.

Bryna Coonin, Coastal Resources Management Librarian from East Carolina University, quickly discussed the findings of her research into the accessibility of approximately one dozen electronic journals with a focus on those used in scientific research. As in other studies, the search screens were targeted for examination, but she also looked at the initial results screen and what kind of output was provided. Bobby was used to discover Priority 1-accessibility errors, which she supplemented by manually checking the pages to ensure a problem truly existed. HTML, better than PDFs for overall accessibility but not for printing, was offered as an option for results display in six of the electronic journals. Image based PDFs can be problematic for almost all but the latest screen readers whereas text-based PDFs are easily understood by the software. Of the resources tested, JSTOR is one provider who exclusively uses image-based PDFs. They are aware of the problem the format creates for the disabled and are working on a solution.

Jennifer L. Ward(jlward@lib.washington.edu ) is a Systems Librarian at the University of Washington Libraries in Seattle, Washington, USA.

Betty Landesman

What is XML?

This is a summary of the all-day preconference sponsored by LITA at the ALA 2000 annual conference in Atlanta on 14 June, 2002 entitled, "What is XML?: How can XML be used in libraries?"

Sara Randall of Endeavor Information Systems began by giving an overview of the topics to be covered in the preconference:

• •

  introduction;

• •

  XML tutorial;

• •

  representing descriptive data using XML;

• •

  data exchange using XML;

• •

  case studies of libraries and XML.

Randall then gave the first presentation, entitled "What is XML and why should you care?" She introduced the concepts and definitions that other speakers would cover in more detail later in the day. XML is:

• •

  an extensible markup language;

• •

  a way to format data so machines can process it;

• •

  intended to be both machine and human readable;

• •

  an international standard accepted by the computer industry.

Libraries have a long history of exchanging machine-readable cataloging data. Randall postulated that because of our lengthy experience using MARC, we tended to use the hammer/nail approach to library standards: "If all you have is a hammer, everything looks like a nail". She clarified that MARC is not a readily extensible standard. There are field and record size limitations. MARC is difficult to use for circulation or invoice transactions, is difficult to read in its native format, and it is not supported outside the library community.

Randall explored the relationships between XML and other markup languages. She explained that XML is a set of rules used to define the structure and content, and does not include the definitions for the presentation of the data. The presentation of the content is handled through other mechanisms. XML can be used to describe a wide range of data, and its rules can be defined for a local, national or international audience. The same data can be easily output to completely different devices and displays, for example an HTML-based OPAC and an e-book reader.

John Kolman of Elissar Software then provided a detailed tutorial that spanned two sessions, pre- and post-lunch. He made it clear that, even with this extended time, he could not cover everything in depth, but would give the audience an idea of where they might need to do more research.

Kolman defined the basic elements of an XML document and the rules and guidelines for constructing and using them. There is a family of elements, including parents, children, and siblings. Once constructed, an XML document can be displayed in a Web browser or formatted through a style sheet. Kolman defined some of the technical aspects of the coding and structure of an XML document, such as the display of special characters and the use of element attributes.

He then explained document type definitions (DTDs) to specify the syntax of an XML document. DTDs ensure that an XML document is correct by validating elements and attributes. Kolman described that there are limitations on the DTDs' ability to define data elements precisely and to support non-character data types. As a result of this limitation, schemas were developed to describe the exact content of data, and are now widely used in XML processing.

Kolman talked about the extensible style language and transformation (XSLT). This is a formatting and transformation processor that can accept any XML as input and use it to output any set of character data based on a set of programming rules.

The above represent the main topics which Kolman explained in much greater detail. He walked the audience through the overall process in a very effective step-by-step process that started with a basic XML document and progressively developed it using increasingly complex processes. He knows his material and can present it in a way comprehensible to relative novices, and his humorous delivery was very well received.

David Rudy from Cornell University followed with a presentation on document type definitions/schemas. He reviewed ten DTDs and schemas used in libraries:

1. 1.

   Dublin Core;

2. 2.

   MARC in XML;

3. 3.

   Metadata Object Description Schema (MODS), a language-based MARC subset which has been described as "MARC for dummies";

4. 4.

   Generic Descriptive Metadata (GDM), a richer descriptive metadata standard developed at Berkeley;

5. 5.

   Visual Resources Association (VRA), a 17-element data set;

6. 6.

   Content Standard for Digital Geospatial Metadata (CSDGM), developed by the Federal Geospatial Data Committee;

7. 7.

   Text Encoding Initiative (TEI), primarily designed to encode data that already exists, e.g. Moby Dick;

8. 8.

   Encoded Archival Description (EAD), used for archival finding aids;

9. 9.

   Metadata Encoding and Transmission Standard (METS), an XML framework or container for metadata to manage and exchange digital objects; and

10. 10.

    Open Archives Initiative (OAI), which does metadata harvesting.

Ralph LeVan, a very lively and amusing speaker, from the OCLC Office of Research's presentation was entitled "Z39.50 as a Web service". He reviewed the weaknesses and strengths of "classic" Z39.50. He stated that Z39.50 is not popular with the Web community because of its connection-based sessions, its binary encoding (which means that a coder cannot type it directly into an editor), the fact that it is transmitted directly over TCP/IP using open sockets and is very complicated. The strengths of Z39.50 include: the result sets are enabled by statefulness, abstraction (which allows for searching by a concept such as "title", regardless of exact field names), and its "explain" function.

LeVan then described OCLC's development of Search and Retrieve on the Web (SRW) and Search and Retrieve with URLs (SRU). SRW is based on Simple Object Access Protocol (SOAP), a wrapper that allows data to be "stuck" into XML. It is semantically equivalent to "classic" Z39.50 without the "baggage" of the standard. SRU is SRW without the SOAP wrapper, and is intended for thin clients.

LeVan proposed the question: Is this a solution looking for a problem? Should libraries stop using Z39.50 and adopt SOAP since it is much simpler and easier to use? LeVan surmises that it is not probable that SOAP will be quickly adopted by the library community, when Z39.50 which does the same thing works well for the library community. LeVan predicts that SOAP will be used first in library communities that deal with geospatial and biodiversity data. In the future as libraries need to interact with other communities, XML will provide the framework for exchanging data.

The final speaker was Roy Tennant from the California Digital Library. He described a number of real-life projects in which libraries are using XML:

• •

  An ILL system at Oregon State University takes information from multiple sources (e.g. OCLC and the local catalog) and creates a single output with all the elements required.

• •

  Service Tasmania is a portal to government services in which an XML metadata repository allows the presentation of different views for different communities.

• •

  The Scholarly Publishing Office of the University of Michigan wanted to develop low-cost, scalable mechanisms for the electronic publication of journals and monographs, and can migrate a word processing document with paragraph styles to XML.

• •

  UCSF has developed a portal for accessing millions of documents from a variety of sources relating to tobacco.

• •

  The California Digital Library is publishing books online, marking up the texts in XML and serving up the presentation(s) dynamically using XSLT stylesheets.

These and other projects are described in a forthcoming book Tennant has written. Tennant's presentation included a number of diagrams and flowcharts in the slides that described in detail the various implementations. He accomplished his main purpose admirably: to demonstrate that libraries are using XML today to do useful work and to accomplish tasks that were heretofore very difficult or impossible to do. Tennant not only knows his material, but he is also a dynamic and enthusiastic speaker; when he told the audience that "You, too, can do it!", he had made it believable.

On the whole, the preconference fulfilled the promise to impart the information suggested not only by its title, which has been used fairly often over the past few years, but more importantly by its subtitle: "How can XML be used in libraries?"

Betty Landesman(bettyindc@yahoo.com) is a Catalog Librarian at the Congressional Research Service in Washington, DC, USA.

Stacy Voeller

Can Electronic Preservation Work?

This is a summary of the LITA's President Program "Can Electronic Preservation Work?" Flo Wilson, the LITA President, introduced the session by stating that more and more, our cultural heritage will migrate to the electronic format. Therefore, collaboration between organizations is imperative to ensure ease of usability between disciplines. Her president's program focused on the successful initiative undertaken by the National Initiative for a Networked Cultural Heritage (NINCH).

The NINCH "was created to help shape a digital environment through intensive, collaborative discussion and thoughtful action of its constituent members". David Green, Executive Director of NINCH, discussed using NINCH as an example of the creation of a globally distributed and accessible collection promoting a networked cultural heritage which shares ideas, resources, experiences and research. The organization strives to develop standards and best practices by providing new approaches and new scholarly tools to enhance the digital environment.

NINCH has also fostered the development of managing knowledge resources responsibly by providing collaborative space, and has moved the process from not just thinking about digital preservation, but actually doing it. Bernard Reilly, Jr, President, Center for Research Libraries, discussed the opportunities for collaboration within the digital arena, and the responsibility of those being entrusted with the information.

Virginia Kerr, Digital Technology Librarian, Northwestern University Library, relayed lessons she learned from participating in the NINCH Building Blocks Workshop on models for effective collaboration in digital initiative development. The implications of computing and how it can better serve the humanities was discussed during the workshop and helped attendees identify possible problems, develop solutions, and explain how to incorporate the digitization process. Kerr described a method of "thinking out loud" discussed during the workshop which enabled participants to work smarter while thinking archivally about existing collections embarking on the digitization path.

Web pages providing examples of digitization projects provided by Kerr include:

• •

  Digital Imagery for Works of Art from Harvard University Art Museums: http://www.dli2.nsf.gov/mellon/report.html

• •

  Digital Imaging & Media Technology Initiative from the University of Illinois Library: http://images.library.uiuc.edu/

• •

  Andreas Vesaluis' Renaissance anatomical atlas On the Fabric of the Human Body from Northwestern University: http://vesalius.northwestern.edu/

• •

  Edward S. Curtis's The North American Indian from the Library of Congress: http://memoryloc.gov/ammem/award98/ienhtml/curthome.html

Museums are attempting to narrow the digital divide by entering the electronic preservation arena as well. Richard Rinehart, Direct of Digital Media, UC-Berkeley and Project Manager for two museum consortium projects, discussed the collaboration efforts California has made in a state-wide initiative for museums and the Online Archive of California (MOAC). As a participant in this project, he learned that in general the museum environment is generally characterized as not being known for their descriptive cataloging of materials, or for their automation efforts.

The findings from the MOAC project demonstrated that museums lack the resources necessary for digitizing their collections. Consistently, most participants were resource-poor, had no in-house specialist for cataloging, and lacked the computer software for creating and gathering the metadata. In order to effectively work together and maintain cohesiveness, it was necessary to develop some best practices based on principles from NINCH. Two sets of best practices were necessary due to the variations of cataloging required by libraries and museums.

The questions and comments from the audience included whether the picture of the object is an acceptable alternative to the actual physical object. Rinehart explained the benefits of making the picture universally available via the Web, and discussed the affects of reproduction on the Mona Lisa. The ability to reproduce this painting has actually assisted greatly with the physical preservation of the painting as more money has been put toward its upkeep than towards any other object.

Another comment from the audience revolved around this attendee's own work with digital preservation with museums and the difficulty she has faced due to the lack of funding for such endeavors. She expressed that more dollars need to be geared towards these types of projects and collaborations.

Stacy L. Voeller(voeller@mnstate.edu) is an Electronic Resources Librarian at Minnesota State University Moorhead, Moorhead, Minnesota, USA.

Susan Thompson

Recruiting and Retaining Library Technology Staff

The following is a report from the program sponsored by LITA's Heads of Library Technology Interest Group entitled, "How Much is Not Enough?: Recruiting and Retaining Library Technology Staff." The ever-changing status of economics and culture makes the search for qualified IT staff a "boom or bust" roller coaster. Yet even in flush times, maintaining an adequate level of technology staffing presents a particular challenge for libraries and information centers. This panel discussion looked into the issues libraries face in developing, recruiting, and retaining a capable IT staff. This was the first program for the LITA Heads of Library Technology Interest Group which was established in 2001 to provide a forum and support network for those individuals with administrative responsibility for computing and technology in a library setting.

Several common themes emerged across the panelist's talks. Libraries cannot compete with the private sector on salary but money is not the only reason people choose to stay in a position. An interesting job in a place the employee values and that values the employee can be as attractive to IT staff as it is to any other type of staff. Features of IT jobs in libraries that are competitive with commercial jobs include training opportunities, flexible job schedules, and varied work assignments.

Do not be afraid of short-term retention. Between the nature of technology jobs and the fact that libraries often have few advancement opportunities for technical staff, it is likely that the IT staff will only stay a few years. However, with proper management, the library can benefit from infusing new blood into the IT department on a regular basis without losing expertise or knowledge.

Finally, it is important to pay attention to the needs of current staff and go out of your way to encourage them to stay. Continually rehire your best people.

Jim Barrentine from the company Information Partners – Technology and Management Solutions for Libraries' works with public libraries and consortia on technology planning including hiring and evaluating IT staff. His presentation entitled, "Staffing library IT departments: finding, keeping, training and maintaining," offered advice on finding, keeping, training, and managing technology staff.

Finding IT staff

First determine what skills are needed. A different set of skills is needed to design a piece of technology compared to operating it, a subtle but important distinction. Most modern library applications are "shrink wrapped" – they are ready to operate upon installation. Therefore, the library needs to decide if they really want to get into the "development business" or if they want to look for other, more support-related skills in their IT staff.

In hiring an IT manager it is important to be able to offer a competitive salary. It is even possible they might be paid more than the library director for whom they work. An IT manager does not need to have an MLS. They simply support the library much as any other technician. If a library is hiring an IT manager and no one on the staff has similar credentials it is important to get help in evaluating the candidate's skills using local computer center expertise or a consultant. Expect to provide management training for your new hire. Since even the biggest library IT department is small compared to traditional IT jobs, it is not likely that a library will be able to attract a candidate with previous management experience.

It is often a good idea to develop home grown IT staff in order to take advantage of technically competent employees already working in the library. Such employees often have the advantage of already being acclimated to the library world and appreciative of the library's mission – important retention factors for any type of employee. Along the same lines, look for people with technical credentials but no experience. Commercial companies are less likely to hire someone with just book learning. Expect to lose these people after a few years as they gain the experience they need to move into private industry. The library will survive just fine if the job is organized to facilitate transitioning to new people. The military has survived quite well with short-term commitments from their "employees."

Properly supervised students, including high school students, can also be wonderful IT employees. However, be careful with ex-dot com employees. They may resent the dramatically lower salary and loss of benefits so they are likely to leave as quickly as they can to return to the private sector.

Keeping IT staff

Libraries cannot compete on the basis of money. However, libraries can compete in other areas. Since many libraries have small IT departments, the IT staff is expected to be a jack of all trades. Many technical people thrive on job variety and may prefer it over the highly specialized positions usually found in larger private sector IT departments. Teach IT staff about the library – even have them work at a library function for a time to give them an understanding of our environment and help them feel a part of the library world.

Other job characteristics libraries can offer that techies find rewarding include continuing education and flexible work schedules. A technical person working on a well-defined project with little people interaction does not need to work traditional hours. Give the IT people visibility – praise them publicly and provide memos or certificates of congratulation. Finally, give them "toys." Techies can find it very rewarding to be allowed to play with new technology, something they often cannot do in the commercial sector. For instance, allow them to experiment with something like Lynix or MySQL on an old computer the library is not using.

Training

"The kind of staff you want always wants to learn new things." Do not hire anyone that does not express enthusiasm for continuing education opportunities. Sources of training include computer-based training, community colleges, and professional organizations such as LITA. Certification in such areas as Microsoft Network Engineer is highly desired and is often available at places like community colleges that may not be that expensive for libraries to use.

Managing IT staff

Someone in the library needs to make sure they understand all the ramifications of developing and managing the library's own IT systems. Documentation is crucial. It should be required that IT professionals document their projects even though they often claim you only need to "read the code."

Use the "two-man rule." Do not do anything unless more than one person understands how to do it. If an IT staff member wants to undertake a project no one else understands, it should be their responsibility to train a backup before having permission to continue. Both documentation and the two-man rule provide the library with a backup for an IT person's knowledge in case they leave.

Finally, consider implementing an annual audit. The audit involves someone on the outside looking at what you do and verifies if it is industry standard. They can also point out good ideas other libraries have implemented that you could benefit from.

The next presentation was by Michael Piper, director, Detroit Area Library Network (DALNET) entitled, "High retention leadership." DALINET supports academic, public, and special libraries in the Detroit area. Piper has spent his first year at DALINET recruiting personnel. He considers team building and establishing a high-retention work culture as two of his most important leadership responsibilities. He identified two myths, which hinder libraries' recruitment and retention efforts:

1. 1.

   Myth No. 1. The unemployment rate makes it easy to recruit and retain talent. In reality, there is never enough good talent. According to Tolben (2001) Winning the Talent Wars, "Your best people are the ones most likely to leave … because they can."

2. 2.

   Myth No. 2. Most factors that influence retention are outside of your control. In reality you can exert far more influence than many libraries realize.

Why do people stay? According to Kaye and Jordan-Evans (1999) in their book Love 'Em or Lose 'Em, 90 per cent of their respondents listed at least one of the top three items as a reason for staying:

1. 1.

   career growth, learning and development;

2. 2.

   exciting, challenging work; and

3. 3.

   meaningful work in which they can make a difference.

Significantly, money appears down in eleventh position on their list.

Key to retention is the employee's relationship to the boss. It is important to think of your colleagues as free agents whom you need to convince to stay. The boss should directly influence his or her star employees to stay and help mismatched employees to find more compatible situations.

There are three steps to being a high retention leader. First is to be the employer of choice. Turnover is expensive in morale, training, etc. – typically costing between 70-200 per cent of a workers salary. IT workers cost even more to replace.

Checklist to be an employer of choice:

• •

  offer chance to learn and grow;

• •

  provide meaningful and challenging work;

• •

  recognize and reward achievement and talent.

Second, take the time to hire the best candidate. A thoughtful recruitment campaign which matches potential employee's values, drives, motivations to those of the organization is the best retention strategy. Therefore, take the time to do it right.

Checklist:

• •

  Differentiate your job opening from other employer's. Look for "wow" factors and play to your organization's strengths. Cite tuition assistance, job security, flexible schedules, and other factors which may not be as common in the private sector.

• •

  Know what you want.

• •

  Treat your applicants like customers.

• •

  Hire for fit, then train for the job.

Finally, build a high retention culture. Continuously re-recruit your current talent. Focus on building relationships from day one. Mentor new employees by providing them with straight answers and insider information on the organizational culture.

Checklist:

• •

  Build-in continuous learning.

• •

  Hold managers accountable. Assess their ability to retain employees, particularly their ability to motivate, care about, remove barriers and share ownership with their employees.

• •

  Seek to build an exceptional culture. The workplace has replaced the local neighborhood. People spend more time with their colleagues.

• •

  Engage souls. People like to make a contribution and feel like they are part of something larger than themselves. Also encourage people to treat each other with respect.

• •

  Articulate a clear sense of purpose and then seek alignment with your employees.

• •

  Focus on taking care of internal customers – your colleagues. Devote time on staff retention and morale. Ask staff, "What would keep you here?" "What would make your work life easier?" "What are your career goals, training needs, etc.?" Make note of every response and then ask yourself regularly, "What have I done this month to meet this employee's needs?" This should be an ongoing dialogue. When in doubt, ask the employee what they need.

Reginald Harris is the head of the Information Technology Support Department at the Enoch Pratt Free Library, Baltimore. His presentation was entitled, "Recruiting and retaining library technology staff: a view from the 'real world'." Reginald is a "home-grown" expert. He often serves as a translator since Systems and the rest of the library do not always speak the same language. He has found that no matter how good the times, there always seem to be a shortage of money and staff in IT.

Pratt Free Library has encountered several IT recruitment challenges. Money is an issue. Pratt is better than many libraries but there is still not enough money to effectively compete with private sector IT jobs. They also share several other challenges that are specific to libraries. Many IT people see libraries as a job "out in the boondocks" – they are not part of the mainstream IT career path. Gatekeepers often block access to IT jobs in libraries. Specifically, why do library IT staff need an MLS? Hospitals and law offices do not require that IT staff be specialists in their areas. This leads to another issue – where can IT go from here? IT staff often cannot move up in the library hierarchy without an MLS.

On the other hand, libraries do have some advantages in IT recruitment. Autonomy is often possible in libraries where IT staff can take control of projects with much less micro-management than in the private sector. Library IT offers a variety of technical challenges with a mix of different systems and jobs to be done. Libraries are a good entry port into the IT industry. Libraries should consider advertising positions to IT graduates just as we advertise for library school graduates. Support is often available from above, the library director, to fund technology adequately and green light projects. It is also highly desirable to have support from below – a more technically savvy library staff will send fewer problems to IT to deal with.

Harris summarized the key to retention – "money and training will keep your people." Libraries may have no money but they can often facilitate training and certification opportunities. While training can be a risk because it makes IT staff more hirable by the private sector, mirroring jobs (the two-man rule) trains a backup on each project so knowledge is not lost when someone leaves.

Other factors that encourage IT staff to stay in library jobs include flexible work schedules. IT people have families too. Unlike private industry, usually it is not regularly expected to work long hours in libraries. Libraries can often offer desirable benefits such as insurance and retirement. Many people value the altruistic goals and values found in libraries. Finally, respect and its intangibles are important including thank yous and acknowledgement at staff meetings. Remember to notice when things are going well, not just when there are problems.

Now that we know what attracts IT staff to stay in libraries, why do they leave? Not enough challenges, the work is not interesting. Not enough support from either above or below. There is a bit of us vs. them in libraries. Library staff sees "them" as coming to do something to their computer. This is where Harris's role as translator is particularly helpful.

Job cycle is an important factor to consider in IT retention. What does retention mean? It is common in libraries to retire after 20 or more years in the same institution or even the same job. Long job tenure is no longer common in the general job market and particularly in IT. IT positions at Pratt have a typical two-year job cycle. Is this really bad when we consider replacing our equipment every three years as desirable? A shorter job cycle may not be a bad thing. It brings in new blood and new ideas. IT people do not like the "we've always done it this way" attitude that may become entrenched in long-term employees.

Harris offered several suggestions on how to recruit and retain IT staff:

• •

  Ask "why would great people want to work for you?"

• •

  Grow your own. Provide technologically promising people with training and move into IT areas. However, there is often a limit on how far the "grow your own" people can go as far as their understanding of the technology.

• •

  Reduce IT staff stress by providing support. Better trained library staff reduces the load on IT. Pratt has a PC patrol of trained library staff who can do triage or even simple fixes on technology problems. All library staff should become somewhat techie. This can be done by encouraging a culture that is not afraid to make mistakes. Failure should be seen as a learning opportunity. In this way staff can come to feel it is OK to take a risk. From time-to-time throw out the rule book.

• •

  Rehire your best people. We already do entrance and exit interviews. We should also do "stay" interviews – what would make you stay?

• •

  Do not expect to run tech departments like a library branch. If a position becomes vacant in a library branch, the manager can usually step in to man the desk until a replacement arrives. If a tech position becomes vacant, the IT manager usually cannot simply take over. While he/she may know the overall concepts, the details are ever changing. Again documentation and the two-man rule help fill the breach.

• •

  Treat technology as a second language – "Techies from Mars, librarians from Venus." Techies want libraries to be less insular, to look outside of the library world for ideas. They see libraries as part of the information delivery business so we should look at what others, like Yahoo, are doing. Look at what is on the techie's reading list. Fortune, etc. with articles like customer service. Dilbert important – techies implore us not to become the pointy headed boss.

ReferencesKaye, B. and Jordan-Evans, S. (1999), Love 'Em or Lose 'Em, Berrett-Koehler, San Francisco, CA.Tolben, B. (2001), Winning the Talent Wars, Nicholas Brearley, London.

Susan M. Thompson(sthompsn@csusm.edu) is the Library Systems Coordinator at California State University San Marcos, San Marcos, California, USA.

Megan Kinney

Major Open Source Web Finding Tools

The following is a report from the program entitled, "Major Open Source Web Finding Tools: Digital Library Systems for Librarians," which was jointly sponsored by the LITA Open Source Systems and Internet Resources Interest Groups. The program opened with the question, "Why use open source and particularly, why for libraries?" A few of the reasons expressed in favor of using open source software were:

• •

  Open source software produces stable systems that you do not have to reboot at least once a day.

• •

  Open source is also highly customizable, yet affordable. For example, if you want a SQL database but do not want to spend $50,000 for a license … there is MySQL (http://www.mysql.com).

• •

  Since open source software is not proprietary, you do not have to rely on upgrades, support, bug fixes, and new features from a big corporation, and you are not tied to one platform or operating system.

• •

  For building digital collections, it is the only realistic solution to long-term access because the code is always easily accessible.

• •

  The most important reason for libraries to implement open source systems is that they promote free and open access to information, an important tenet of the library as an institution in the USA.

Next there were two demonstrations of successful Web finding tools built with open source software, Greenstone and INFOMINE.

1. 1.

   Greenstone by the New Zealand Digital Library Project. Greenstone (http://www.nzdl.org) builds robust, individualized, multimedia, searchable, Web based digital library collections. Some of the notable features/functions include: the ability to expand and collapse searches, as well as extract phrase hierarchies, full-text search, subject browse (hierarchy), browse by organization, thesaurus browse. The interface is comprised of "bookshelves," a nice way to introduce new technology through a familiar analogy. Some of the collections include: oral history (voice and pictures), melody index (synthesized .gif image with the same notation as the music itself), Chinese collection (with both English and Chinese interface), Arabic collection (with both English and Arabic). Sounds great, but it must be hard to build and administer a collection … not with "The Collector," a wizard for building new collections. Please see the Web site (http://www.nzdl.org ) for more information, as well as the forthcoming book: How to Build a Digital Library (Witten and Bainbridge, 2002).

2. 2.

   INFOMINE by the University of California. INFOMINE (http://infomine.ucr.edu) is a virtual library for university level research that provides access to the Internet through a traditional library classification system (LC Subject Headings). In order to appreciate this virtual library, one must see the differences between books and Web sites:

• •

  Books: easily defined, static, permanent, LC has 119 million, limited number of publishers.

• •

  Web sites: hard to define, dynamic, frequently disappear, 2 billion pages (Google), anyone can publish a Web site.

INFOMINE has been successful by simplifying the problem (of the size and scope of the Internet) and making it work using editorial standards, automated assistance, and wide collaboration. Partners include: Wake Forest University, California State University, the University of Detroit – Mercy, and others. The main method for browsing and searching INFOMINE is by category (based on LCSH and each with its own editor).

ReferencesWitten, I.H. and Bainbridge, D. (2002), How to Build a Digital Library, 1st ed., Morgan Kaufmann, New York, NY.

Megan C. Kinney (kinneymc@umich.edu) is the National YOUTH Access Coordinator for Libraries for the Future in Phoenix, Arizona, USA.

Jonathan Rothman

Let's Not Forget about Tomorrow

This is a review of "Preserving Access to Digital Collections," a program presented by the LITA (Library Information Technology Association) Technical Issues in Digital Data Interest Group. This session was moderated by Jennifer Weintraub, co-chair of the Technical Issues in Digital Data Interest Group who began by informing the audience that the IG has been renamed as the Digital Technologies Interest Group. This was a very full program that included information-packed presentations from three interesting speakers.

Vicky Reich the Director of the LOCKSS Project at Stanford University began with a description of the Lots of Copies Keep Stuff Safe (LOCKSS) Project. The goal of this project is to build a cooperative, decentralized, digital archiving system based on the existence of many copies of each electronic file. The initial system is designed to archive electronic journals, but it is expected to broaden to encompass other materials. The basic approach is inspired by the way libraries have traditionally handled paper materials – librarians protect materials from becoming "unpublished" by collecting many copies. Ms Reich noted that, "… even if you were a mean person, it would be very difficult to destroy all of the copies …"

With LOCKSS, libraries install caches that hold copies of the content that they are authorized to access. These caches are never flushed. The cache management software crawls through publishers' sites, downloading copies of authorized content as it is published. Patrons access the content via a special proxy server that normally connects them to the publishers' sites. In the event that a publisher's site is unavailable, the proxy automatically redirects the patrons to the content stored in the local cache. In addition, each cache knows about some (but not all) of the other caches and the caches continuously audit themselves against each other. The software will automatically attempt to obtain missing content when discrepancies are noted and can also perform protective functions such as issuing warnings when the number of known copies of any content file drops below a threshold. The primary new thing about LOCKSS is the protocol by which the caches talk to each other.

Libraries take over a good deal of responsibility from publishers under this scheme, but the only power that publishers give up is the right to revoke access to content once is has been granted. Funding for the project has come from the Mellon Foundation, the National Science Foundation, Sun Microsystems, and a private donor. Currently there is a beta test in progress involving caches at over 50 libraries worldwide. The beta test is scheduled to end in August 2002 and production rollout is predicted for 2004-2005. For more information, see http://lockss.stanford.edu.

Martin Halbert the Director of Library Systems at Emory University Library spoke on "The meta scholar initiative and facilitation of access to digital collections." He began by noting that he has heard from archivists that "Digitization inverts the preservation equation." He went on to explain that, with print, increased discovery and usage tends to reduce the likelihood of collection survival due to physical wear and tear. With digital materials, however, the opposite is true. Materials that are not discovered and used are less likely to survive. Emory has undertaken several studies of preservation practices for digital collections nationwide and has concluded that we, as a profession, are doing a "… woefully inadequate job." There are large, widespread problems that will not be solved quickly, but standardized metadata does offer a start, and OAI is the most promising option for standardization.

As a response to these issues, Emory is spearheading the metaScholar Initiative (www.metaScholar.org), a series of projects designed to explore and expose issues in using OAI. It is based at Emory, includes many partners, and has received funding from the Mellon Foundation. There are two primary projects currently underway:

1. 1.

   The metaArchive Initiative (www.metaArchive.org) is a cross-institutional scholarly metadata harvesting and searching service. This service will aggregate metadata from multiple institutions and will provide a searchable Web interface designed to provide coherence across interdisciplinary subject domains. The metaArchive Initiative will also provide metadata conversion services through a suite of open source software designed to make OAI a practical alternative for smaller institutions.

2. 2.

   American South (www.american South.org) is an OAI-based aggregation network providing a collaborative digital collection of Southern History and Culture. The primary focus of this project is on how scholars will use the collection. Work to date has focused on the harvesting network and a Scholars' Portal to provide appropriate access.

Adrienne M. Woods, Archivist from the National Archives and Records Administration (NARA) presented a talk entitled "Building the archives of the future: NARA's electronic records archives program". She began by explaining NARA's dilemma. Since 1935, when NARA was founded, its job has been preserving paper records. Now that we have moved to e-government and e-commerce "… electronic records are critical for the effective functioning of democracy." Some of the record types NARA must cope with include: e-mail, geo-spatial data, office automation products, databases, images, and complex documents. The volumes involved are huge: the Clinton Administration in just one of their many e-mail systems generated 40 million e-mails; the State Department generates about 25 million e-diplomatic messages a year; the Department of Defense generates 54 million e-official military personnel files annually; and the Census Bureau generated 600-800 million electronic images from the 2000 census. NARA is responsible for preserving all of these, but there is no single system that can handle all of the varying types of digital records at this point. NARA's responsibility includes not just preservation, but making the data accessible. They have been preserving electronic records for 30 years, but the holdings have been relatively simple files. Ms Woods summed it up: "We can save anything, but whether you can access it …"

With the Electronic Records Archives (ERA) program NARA is looking for new methods of electronic preservation. Basic requirements are: persistence ("Eighteen months won't do it … the data have to last until the end of the Republic …"); authenticity – it must be possible to verify that the data has not been changed; and scalability. Conceptually ERA will be based on OAIS. While the goal of ERA is a system that will meet NARA's requirements, it is being specified as a framework that could be used for other systems once it is developed. The project timeline is:

• •

  2002 – begin R & D, Finalize Key Documents, Staff Project Management Office.

• •

  2003-2007 – continue R & D, Develop and test online public Access, Hire Systems Developer.

For more information on ERA, see www.nara.gov/era

The audience responded to the speakers with significant interest. Due to time constraints, only a few questions were possible and the session ended with a number of informal discussions in progress between the presenters and audience members.

Jonathan E. Rothman (jrothman@umich.edu) is a Senior Systems Librarian/Analyst at the University of Michigan University Library, Ann Arbor, Michigan, USA.

One should not forget the popular and exceeding capacity attendance of the Top Technology Trends session that LITA hosts at each Midwinter and Annual ALA conference. Reports and coverage of that important session are posted and archived on the LITA Web site at: http://www.lita.org/committe/toptech/mainpage.htm At the time this issue went to press reports of the ALA 2002 conference were yet to be released.