Spring Conference -- Something Old, Something New: Looking at Standards
4/12/2002
Hogan Campus
College of the Holy Cross
Worcester, MA
Morning speaker: Sally McCallum, Chief, Network Development and MARC Standards Office, Library of Congress:
The Metadata and Markup Explosion
Breakout sessions:
Sam Dempsey (Informata.com): ONIX In-Depth
Norman Desmarais (Providence College): XML Basics
Linda Tiernan Kepner (Peterborough Town Library): Workable Solutions with Dewey Classification
Robin Wendler (Harvard University Library): Adopting and Adapting Metadata Standards
Lunchtime speaker: Birdie MacLennan (Bailey/Howe Library, University of Vermont):
Highlighting NASIG: Serials and Beyond
Afternoon speaker: Lynne Howarth (Faculty of Information Studies, University of Toronto):
Making sense of metadata: towards a common core for bibliographic control
Sally McCallum, Chief, Network Development and MARC Standards Office, Library of Congress
The Metadata and Markup Explosion
It is not unusual nowadays to encounter, at conferences and in the literature, overviews of the proliferation of metadata standards which now exist to serve diverse communities of interest, within the library world and well beyond. What is less common, at least in this writer's experience, is for a speaker to not only survey the status quo in all its multiplicity, but also to clearly delineate actual recent trends (as compared with predictions based on extrapolation) and provide a look at concrete, practical developments which have promise to address the increasingly-voiced desire for interoperability. The title of Sally McCallum's talk might lead one to expect another of the aforementioned surveys, but in fact she provided her listeners with a sense of some ways ahead, post-explosion.
Ms. McCallum began with a review of where we are presently, including the pressures on long-established practices. The three primary topics of interest are 1) content, or the data itself, such as MARC cataloging data; 2) markup, or data tagging, such as MARC 21 tags and subfield codes; and 3) the structure given the data, including (again in MARC terms) the leader, directory, etc. For the most part, we have since 1970 operated in a remarkably stable environment with regard to these three areas. Our primary standards, for example the MARC 21 formats or AACR2, have adapted through evolution, allowing for stability and development simultaneously. However, a number of pressures and complaints have more recently put these standards, and this developmental pace, to more serious tests.
Chief among the pressures are the enormous numbers of electronic and composite resources, which are highly mutable and present serious problems of version control; the challenges of "repository management" for electronic collections; expanded user expectations regarding retrieval, including the desire for cross-collection retrieval; and the rise of new computer language technologies, with XML being a prime example. Notable complaints include the intentional duplication of data in cataloging record (although the reasons for this are not always understood); the belief that there are "too many" data elements in MARC 21 (although the format serves many communities and purposes); skepticism about the possible overparsing of data (although movements for less parsing are consistently opposed); and questioning of the need for or utility of ISBD punctuation, as being of debatable functionality online (although this also has its adherents).
Ms. McCallum then addressed what is "exploding", and what is not, with regard to the three primary topics of interest. Content is exploding in terms of multiple forms of descriptive data (not only MARC data, but data elements specified in EAD, VRA, ONIX, etc.); in terms of differing sets of rules for description (cf. Dublin Core's "creator" as compared with ONIX's "contributor"); and also in that more types of data, such as administrative, digital geospatial, and rights management data, have come to prominence. But on the other hand, the global library community seems to be converging on AACR as a descriptive standard, with the ISBDs as an important factor in this. Markup appears to be exploding: we have multiple forms of MARC (including UNIMARC), Dublin Core (DC), ONIX, EAD, HTML tag sets and more. A contributing factor here is the relative ease of establishing XML tag sets. But by contrast again, the global library community seems to be converging on MARC 21 and EAD to some extent. There is, finally, an explosion in data structures. ISO2709 is the structure standard on which MARC is based, but there are also multiple whatever-MLs (especially SGML, XML, HTML), proliferating DTDs, schemas, the Microsoft Access database structure, and more. But countering this proliferation, there is again an apparent convergence on XML and schemas in the computer community.
In each case, there is arguably a centralizing tendency that contributes some stability to the metadata and markup explosion. Still, it is critical to develop tools for interoperability among standards, since it is not likely that the various library and information communities will be able (or actually ever need) to settle on single, shared standards for content, markup and structure. The balance of Ms. McCallum's talk was devoted to a "possible ways forward," the MARC Tool Kit (http://www.loc.gov/marcxml) and the Metadata Encoding and Transmission Standard (METS) package (http://www.loc.gov/mets), both available from the Library of Congress web site. In essence, the Tool Kit repackages MARC 21 data into an XML schema, which supports communication of MARC and also supports transformations of the data for special uses. The basic transformation is a lossless, round-trip conversion of the MARC 21 record from its ISO 2709 structure into an XML structure, preserving its highly developed semantic content. The schema is simple and flexible, and will adapt easily to changes in MARC 21. The open-source conversion programs are maintained by the Library and can be downloaded.
Other major elements in the MARC Tool Kit are the DC and MODS transformations, validation software for structural, definitional, and semantic checks of records, and transformations between different tagging styles and character sets. The DC transformation builds on the MARC-DC mapping already maintained by the Library, and again is available as open source. MODS (or Metadata Object Description Schema) is a newer concept, addressing the need for a simpler schema that is nevertheless MARC/AACR-compatible and richer than DC. It is a "friendly" schema without coded values (further information is available at http://www.loc.gov/mods). The Tool Kit will also include software for experimental use that segments components of unit records and brings them together for display purposes, in accordance with the work/expression/manifestation/item model of the Functional Requirements for Bibliographic Records (FRBR).
There are many uses envisioned for the components of the MARC Tool Kit. It standardizes MARC data across various communities, for XML communication and manipulation; opens MARC to XML programming tools and presentation style sheets; assists original input to XML systems; and provides a MARC standard for metadata harvesting. It facilitates packaging metadata records with electronic resources themselves, and provides an evolutionary path, retaining MARC's rich semantics through the evolution into the next generation. The Tool Kit should help system developers, as the Library intends to maintain its components. Finally, it encodes descriptive data for inclusion in a Metadata Encoding and Transmission Standard (METS) package.
METS addresses the explosion of types of metadata, allowing not only descriptive, but also administrative and structural metadata to be included in a single XML document that accompanies an electronic resource. It is a recent initiative of the library community and the Library of Congress has become the maintenance agency for the METS schema. A typical METS package consists of a structural map which outlines the hierarchical structure for the digital object, a file group segment which list the files comprising the object, descriptive metadata such as a finding aid and/or bibliographic records describing the object, and administrative metadata (such as rights, provenance, master/derivative relationships, and migration/transformation history). Both descriptive and administrative elements may be included in the METS package or be external and pointed to by the METS package, and the metadata may be XML or other structural types. METS is beginning to be used at the Library of Congress by the Motion Picture, Broadcasting, and Recorded Sound Division for digital materials. It will have important uses for repository management, and complex document interchange, as well as enabling resource retrieval, object validation, and rights management, among other uses.
To conclude, Sally McCallum reviewed the main points of her talk. Developing flexible techniques for packaging descriptive metadata and incorporating diverse metadata types are essential for staying abreast of the "explosion." XML, as a new record syntax, promises much in this area. The MARC Tool Kit and METS package are initiatives that should help bring this promise to fruition.
Reported by David Miller
Curry College, Milton, Mass.
NETSL Writer/Editor
ONIX In-Depth
Sam Dempsey (Informata.com)
ONIX (an acronym for ONline Information eXchange), uses XML to give publishers a standard format to provide information about books and book-related products. Sam Dempsey, Director of Data Development at Informata.com, gave an informative and entertaining talk on ONIX, providing the participants with a brief history of its creation, how it is helping publishers make their presence felt on the Internet, and how it can benefit libraries as the use of this standard spreads throughout the industry.
ONIX was the result of a joint effort on the parts of organizations in the United States, United Kingdom and Europe. The latest version of ONIX was released on July 25, 2001.
As the Internet has become more and more important to publishers as an outlet to advertise and sell their books, the industry has come to realize that the quality of the bibliographic data and digital content displayed on the Web greatly affects the volume of sales of a title. Now book publishers can use this standard to describe their books in a format that can include not only the basic information, such as title, author, subject and audience. ONIX allows a publisher to include digital content or links which can show book covers, descriptions, full length reviews, even audio and video clips.
Dempsey explained that in the past, the quality of a publisher's description of a title varied widely. He said he had even received information on cocktail napkins for inclusion in a vendor's catalog. ONIX's potential efficiencies for publishers have motivated some to forego their more traditional methods, especially when they can see that greater sales of their titles is the result, he said.
Several of the problems he has encountered drew some laughs from the audience, especially when he was describing how publishers "don't use the same name for an author every time." At that point he lamented, "We really should create some sort of file that would have the author's name and all the variations." Sally McCallum from the Network Development and MARC Standards Office at the Library of Congress, who was attending the session, replied "I know where there is just such a file. Perhaps you'd like to buy it?"
The benefits of ONIX for libraries are "downstream," as Dempsey explained. As more publishers and vendors use this standard, it gives libraries more accurate and timely information on books they are contemplating purchasing. Also, as both library OPACs and their users become more sophisticated and demanding of graphic images and better information, eventually they will be able to download pre-order records from vendors with much more information, and have access to images, descriptions, tables of contents, reviews and excerpts, which will benefit users of the OPAC.
Plans are also in the works to expand the reach of ONIX. Representatives from the Latin American book community are participating in the international development of the ONIX standard and beginning to implement locally. A version of ONIX for serials is also planned for release shortly, he said, as well as one for video/DVD. New releases of ONIX come out about every 6 months and are backwards compatible; version 2.1 is scheduled for release in August 2002.
In his handouts, Dempsey included several websites that interested persons could visit to find out more about ONIX: see http://www.editeur.org and http://www.bisg.org.
Reported by Janet Hughes
Student, Southern Connecticut State University, Department of Information and Library Science
XML Basics
Norman Desmarais (Providence College)
Norman Desmarais, Acquisitions Librarian, Phillips Memorial Library, Providence College, gave an overview of XML, its relationship to SGML and HTML, and offered possible uses for librarians. For a further discussion of this topic, one is encouraged to read Mr. Desmarais' book The ABCs of XML: The Librarian's Guide to the eXtensible Markup Language (New Technology Press, 2000).
Mr. Desmarais defines XML, which stands for eXtensible Markup Language, as a "meta language, a language for describing languages - analogous to metadata, which is data describing other data". Initially created for use on the World Wide Web, XML can be used for any type of electronic publication. As compared with SGML (Standard Generalized Markup Language), an international standard that describes generalized structural information - i.e. the logical and physical structure of documents, or how a document should be laid out and structured - XML uses the principles of SGML to describe the information content of a document.
The developers of XML intended it to be easily transferred over a network; to be capable of supporting a wide variety of applications; to be easy for writing programs which process XML documents; to possess a minimal number of optional features in XML; to make XML documents legible by humans, reasonably clear, easy to create, and consistently processed by the receiver; and to offer a separate markup from content.
Since XML is seen as having all of the above-mentioned facilitating features, one needs to compare them to those of SGML and HTML (HyperText Markup Language). On one hand, SGML, which is considered a flexible tool once it is learned, is nevertheless perceived as complicated, requiring a steep learning curve and high maintenance - for example, it is often necessary to hire a consultant to implement and manage SGML. On the other hand, HTML, basically an application of SGML with a Document Type Definition (a set of rules governing the structure and element content of a document), is also limited since it only deals with layout. In other words, it defines how the information is displayed, such as the colors used or font size of a word in the document; has a limited number of tags; and does not touch upon the meaning of a word. In fact, XML, because of its flexibility in creating tags, or the metatags, which place terms in context and give them meaning, describes the information content of a document more precisely than HTML. The metatags allow for greater functionality and accuracy when retrieving data in the various fields, when accessing unstructured text as well, and generally when creating, manipulating, "reading", presenting and searching data.
An example of the flexibility of XML is shown with a term such as "bill", which offers a variety of meanings depending on the context - personal name, obsolete weapon, a draft of proposed Act of Parliament... to mention a few of the possible meanings. XML provides the syntax that allows users to create their own markup language and define their own vocabularies to meet specific application or industry needs. Not only that, but the XML syntax allows for a common schema document syntax, something of a standardization of the multitude of examples of DTD syntax.
Ironically, it is the very flexibility of this language that creates problems, inasmuch as it is necessary to have consistency when identifying terms. This places the responsibility on each industry or business to find a common way to describe the information in an XML document. The issues, Mr. Desmarais points out, range from deciding what to describe to deciding what to name each field. Furthermore, one should also point out that although XML is relatively simple, it entails becoming familiar with a whole new vocabulary. But to its advantage, almost all XML engines have a "validator", which examines whether the document is well formed and identifies problems to be corrected.
When it comes to possible library uses, it is favorable to have a standard schema for describing and exchanging data regardless of the nature of the data, which system stored it, or how the receiving system will use it. In the library field, MARC can be easily the XML Data Schema for library applications, as it contains all the information needed by librarians or anybody working in the book industry. In fact, when cataloguing, a librarian can distinguish between all different types of author entries or title entries, for example, while, a book dealer or student might be interested in sparser bibliographic information. The use of MARC format as an XML Data Schema would also differentiate between documents by and documents about an author. Needless to say, a unifying schema would allow for a more uniform way of searching, something also achieved by Z39.50 engines, which however require the data to be re-mapped.
The Library of Congress and the British Library have mapped MARC to XML, and developed programs to do the conversion. ONIX is the schema currently being used by the book industry (publishers and vendors) and will probably be enriched with cataloging data from the Library of Congress and the British Library, for use in the library world.
Implementations of XML products in the library may include, to mention only some, the use of the XML linking language (Xlink), the XML Pointer Language (Xpointer), the eXtensible Stylesheet Language (XSL), and XSL Transformations (XSLT). Xlink, for example, allows for a bibliographic link to go from a text to a footnote to a cited article to the text of a book, and back to the starting point without having to retrace one's steps (roundtrip linking), while Xpointer permits one to go directly to a location in the document. Stylesheets allow using the data or document in a variety of ways, without having to alter the original document. For example, bibliographic data can be used for publishers' catalogs, sales promotions, invoices, catalog records, footnotes and bibliographies, just by changing the stylesheet. XML stylesheets are also backward-compatible with cascading stylesheets used in HTML.
Many in the information industry pursue the developments regarding the production and implementation of XML. To name only some of the players mentioned by Mr. Desmarais: Logos Research Systems, Inc., the Graphical Communication Association Research Institute and the Data Interchange Standards Association, Microsoft Corp., IBM, Endeavor Corp., OCLC's SiteSearch, ExLibris, etc. Some sites to visit for more recent developments regarding MARC and XML are: a MARC and XML Tutorial at http://www.bpeters.com, a project of Lane Medical Library at Stanford University at http://xmlmarc.stanford.edu, and Cambridge University Press, http://www.cup.org.
Reported by Ivana Curcic
Student, Simmons College Graduate School of Library and Information Science
Workable Solutions with Dewey Classification
Linda Tiernan Kepner (Peterborough Town Library)
In her hour-long presentation, Linda Tiernan Kepner, Assistant Director/Cataloger for the Peterborough (N.H.) Town Library, was a convincing cheerleader for the Dewey Decimal Classification system. The system was initially developed for a college library and Tiernan Kepner views it as a "truly universal system of classification ... able to encompass all knowledge ... and predict new knowledge." She stated that the Library of Congress Classification System, by contrast, was formulated for congressmen and "does not work well for other subject specialties".
Tiernan Kepner shared humorous anecdotes about Melville Dewey (who spelled his own name Dui): how he started the American Library Association for women to have "another outlet" besides being housewives, and that he was the first to get ousted from that very organization for "financial diddling". Her fond nickname for Dewey - "Cave Man" - brought to mind a different man than most of us are used to reading about.
It would have been interesting to have more time with Tiernan Kepner, so she could have expounded on the problems and "workable solutions" of Dewey, but the hour was amusing and well-spent.
Reported by Cynthia Clifford
Student, Simmons College Graduate School of Library and Information Science
Adopting and Adapting Metadata Standards
Robin Wendler (Harvard University Library)
Robin Wendler is the Metadata Analyst in Office for Information Systems of the Harvard University Library. According to the Harvard LDI homepage, Ms. Wendler "provides expertise about all things metadata to the Harvard University Library Digital Initiative team, to libraries developing LDI proposals, to funded LDI participants, and to the Harvard community at large." Before she launched into her presentation on metadata standards, Ms. Wendler pointed out that the title of her talk was somewhat misleading, because as an employee of Harvard, she cannot really be counted as a representative of the "real world". However, due to the richness of Harvard's collections and the emphasis the library has placed on digital initiatives, she is in a position to test the boundaries of schema as they are developed. Her presentation was a very informative overview of the current metadata standards, with examples of how Harvard University is applying them. For a detailed description of Harvard's Library Digital Initiative, I recommend their Website: www.harvard.edu/ldi.
Ms. Wendler distributed handouts showing examples of geospatial, archival, and visual resource metadata from the Harvard databases. Each example included the record seen by the public and the encoded view, showing the levels of detail that are possible with metadata encoding, and the similarities and differences in coding needed to accommodate a variety of formats. Some descriptive metadata standards can accommodate far more detailed information about the content of files than MARC records, while some are more slender, or simply represent a different way of thinking about the material. Other metadata standards cover technical, structural and administrative information about digital objects.
Harvard has elected a belt-and-suspenders approach to resource discovery. Materials are cataloged at the title or collection level in the ILS. In addition, many kinds of resources receive additional description according to standards specific to their form or topic. Large electronic resources are also listed through a separate World Wide Web portal. While Harvard's ILS is capable of accessing the digital objects in the collection, the portal allows additional metadata descriptions of the digital objects including, for example, more access and use details than one would typically find in a MARC record. Access through the portal also makes it easier to find these important and often expensive resources than it would be in the 9-million-record ILS. Other kinds of resources, digital and traditional, also have access in addition to the ILS. Archival collections, while represented by a collection-level record in the ILS, may also have extremely detailed Encoded Archival Description finding aids in a separate union catalog. Visual images and cultural artifacts are cataloged according to a variant of the Visual Resources Association Core Categories. And geospatial data can also be accessed through a separate catalog that emphasizes the geographic coordinates that they cover. The hope is that whether users begin their search in the broad ILS or in a more focused catalog, pointers among these descriptions will lead the searcher to additional material.
The presentation ended with a list of "Considerations for adapting or creating a metadata scheme" that is very helpful for any library involved in a digitization project. Beginning with analyzing what you want to accomplish, the list outlines such steps as familiarizing oneself with available standards and schema, choice of data structure, and analyzing each subelement for functionality and usefulness. In the final analysis, creators and adapters of metadata schema must consider the following question. "Why are you keeping this information? If you do not know that, you may not be designing it correctly, or you may find that you don't even need it." The sign of a good metadata record, like any other catalogue record, is that it is uniquely descriptive and informative, without extraneous or confusing details.
Reported by Christina Bellinger
Dimand Library, University of New Hampshire
Birdie MacLennan (Bailey/Howe Library, University of Vermont)
Highlighting NASIG: Serials and Beyond
In the second of a series of brief after-lunch talks highlighting the work of related library organizations, Birdie MacLennan described NASIG, the North American Serials Interest Group. Ms. MacLennan, who received the second annual NETSL Award for Excellence in Library Technical Services, is the co-chair of NASIG's Continuing Education Committee.
Founded in 1985, NASIG is an entirely volunteer organization which promotes communication, information, and continuing education about serials and scholarly communication. Its members (currently between 1200-1300) include librarians, vendors, educators, publishers and others who are interested or engaged as participants in the "serials information chain." Core concepts which drive the organization are a "level playing field" uniting members from different backgrounds, teamwork, camaraderie, and respect. NASIG is best known for its annual conferences, usually held on a college campus, its intensive combination of programs and networking opportunities having been described as a "serials spa." The organization also has thirteen standing committees, provides a number of grants and awards, and sponsors continuing education and regional programs on serials-related topics. For example, NASIG has provided sponsorship of a number of speakers at NETSL and New England Library Association conferences.
NASIG is a most welcoming organization with many opportunities for involvement and professional growth. To find out more, you may visit its web site: http://www.nasig.org.
Lynne Howarth (Faculty of Information Studies, University of Toronto)
Making sense of metadata: towards a common core for bibliographic control
The focus of Lynne Howarth's afternoon talk was the work of the IFLA Section on Cataloguing Working Group on the Use of Metadata Schemes, and its likely production of a document which can be used to evaluate the use of existing metadata schemes in given contexts, or to serve as guidelines for the development of new schemes. She observed that we have been dealing with metadata standards for over a century, and asked the audience to consider, with regard to the NETSL program title, if this effort was something old or something new? She also humorously observed that the "transformations" of data discussed by Sally McCallum in her morning talk suggested to her that librarians are now "transformers", like a "modern-day Mr. Potato Head," taking on different shapes and adapting to different situations.
The Working Group (WG) was established at the 1999 IFLA conference in Bangkok. The WG has three "terms of reference," or aspects of its charge. The first is to monitor metadata standards development and applications/implementations internationally. However, at this point there are far too many standards and applications existing for the WG to really pursue this effectively. The second is to determine what the elements of a "core record" for metadata schemes could be. This is the main focus of the WG at present. The third term is "to provide guidance/guidelines on the appropriate use of metadata records and bibliographic records in libraries." Prof. Howarth pointed out that recommendations resulting from this term of reference are not meant to be prescriptive, as the WG does not intend to serve as the "metadata police."
To begin the process of determining the elements of a metadata "core record," the WG began with a paper by Olivia Madison (Iowa State University) which compared Dublin Core elements to the Functional Requirements for Bibliographic Records (FRBR), followed by another by Eeva Murtomaa (Helsinki University Library) which mapped FRBR to the International Standard Bibliographic Description (ISBD). (The former paper is available at http://www.ifla.org/VI/3/icnbs/mado.htm; for the latter, see http://www.ifla.org/IV/ifla64/065-74e.htm.) What was desired was a representation of core metadata elements that addresses both the eight areas of ISBD description and the four generically-defined user functions of FRBR: find, identify, select, and obtain. (An Adobe Acrobat version of FRBR is available at http://www.ifla.org/VII/s13/frbr/frbr.pdf).
Members of the WG were assigned to determine how analogous data elements were represented in each of nine metadata schemes: Encoded Archival Description (EAD), Dublin Core (DC), Government Information Locator Service (GILS), Text Encoding Initiative (TEI), Visual Resources Association (VRA) Visual Document Categories, Consortium for Interchange of Museum Information (CIMI), Digital Geospatial Metadata (DGM), ONIX, and MARC. Crosswalks have been constructed to aid with this comparison (see http://www.fis.utoronto.ca/special/metadata/shortwalk.asp), although, as Prof. Howarth noted, the construction of crosswalks involves as much art as science. Some elements can be mapped across all schemes, others may only have correspondences with one or more other schemes. Additionally, some schemes such as DGM contain elements that could not be mapped to anything else.
Based on this analysis, the WG identified ten core metadata elements are schema independent, that is, which are applicable regardless of domain. They are: subject, date, conditions of use, publisher, name assigned to a resource, relationship to other resources, language/mode of expression, resource identifier, resource type, author/creator. Note that the names of some elements (for example, "name assigned to a resource" rather than "title") need to be broadened beyond their representation in a traditional library context. The WG guidelines will present this "core of cores" (so named by Lois Mai Chan) with a brief definition of each element, the treatment of each element as found in each scheme compared, and any problems or issues associated with an element.
The WG is also investigating the questions of when metadata records from different schemes would be appropriate for different uses, and whether they should be used on their own, in the place of "traditional" bibliographic records, or as a complement to such records. Prof. Howarth showed examples of Canadian web sites that have implemented different types of metadata, with positive results. The Toronto Public Library's Virtual Reference Library added enhanced DC metadata to its descriptions of historical archival materials, and have found increased in-person demand for rare and special materials. The Department of Foreign Affairs and International Trade has a research portal for its resources. These include standard formats such as books and videotapes, but also include transcripts, briefing books, and (nonvirtual) databases. Local records are created for these non-standard materials. A standard metadata form is available for members of the public to create their own records. One outcome of making this form available is that field specialists in different ministerial governments are contributing records, as a means of raising the profiles of their institutions. This raises the profile of the Department's library as well, and contributes to the Department's knowledge base.
A number of challenges remain for the WG. Among others, these include:
The relationship of the "common core" to other descriptive standards or to other metadata standards;
Relating the "common core" more closely to FRBR, both theoretically and in application;
Relating the "common core" to other ongoing work in the metadata community;
Expanding the English-language "common core" to a broader multilingual/script environment; and
Remaining open and flexible to the changing IT environment, to possible new or emerging resource types, etc.
Opportunities for the WG include liaising with other communities (such as W3C) or domains (such as publishing and rights management), exploring opportunities for developing syntactic and semantic interoperability; facilitating university access to a variety of resources for a broad constituency of domains; and bringing traditional skill sets to the still-emerging digital arena. A worldwide review of the draft guidelines will solicit opinion and feedback from interested individuals around the world: all will be welcome to provide input.
For further information on current research in metadata standards and interoperability, see Prof. Howarth's web site, "Modelling a Metadata Ontology," at http://www.fis.utoronto.ca/special/metadata.
Reported by David Miller
Curry College, Milton, Mass.
NETSL Writer/Editor
|
