Start time: 2 pm. Location: SML 409. Present: Matthew Beacom, Tom Bolze, Daniel Lovins (chair and recording), Youn Noh, Britta Santamauro Daniel reviewed the topics he hoped would be covered in today's meeting, and reminded people to check the minutes from last week to make sure their views (e.g., about department names) had been accurately represented. 1. Youn reported on a PREMIS conference and Digital Library Federation (DLF) forum she had attended. She cited a University of Virginia RDF-based metadata aggregator called NINES, and a project at MIT called SIMILE . Daniel asked whether studies have been done on how well RDF worked as a semantic "switching language". (After the meeting, Youn referred him to a SIMILE "Scalability Report"). Matthew pointed out that RDF isn't really a switching language, that it's more like XML, and recommended Eric Miller's article. Youn agreed that it was worth reading. Youn cited a DLF presentation on PennTags, a service whereby members of the University of Pennsylvania community can tag web pages, OPAC records, and other digital objects with the identifying terms of their choice. Daniel mentioned that a presentation on PennTags was made by Laurie Allen at the NELINET OPAC 2.0 conference as well. Allen described how non-librarians have created annotated bibliographies and "shadow collections" out of tagged OPAC records, in some cases providing subject expertise not otherwise available to the library. Tom asked whether it is really a good thing for virtual collections (e.g., "shadow catalogs") to be defined by social tags rather than controlled LCSH-type vocabulary, noting that it seems to contradict our commitment to standards-based cataloging. Daniel suggested that the two might not be mutually exclusive, but rather satisfy different user needs. For example, the phrase "Hurricane Katrina" took several months after the event to appear in LCSH, but began showing up immediately in del.icio.us "tag clouds". At the same time, it seems that user-assigned tags will have brief shelf lives, and therefore should not be regarded as a substitute for subject headings or classification. Matthew pointed out that controlled vocabularies are better at dealing with ambiguity. For example, multiple users assigning the tag "Katrina" to selected digital objects would not all be refering the 2005 hurricane by that name. Daniel suggested that an ideal system would support both tagging and controlled vocabularies. The rapidly growing LibraryThing, for example, supports both. (He mentioned that LibraryThing's Abby Blachly gave a talk at the NELINET conference.) Matthew suggested that the vocabulary of social tagging might not itself be standardized, but that the technical specifications underlying them would be, and that this is what allows the tags to be highly customizable and portable across platforms. Speaking of standards, Matthew reminded us that the SCOPA forum on December 8th will be on "Cataloging Cultural Objects," and that he's one of the featured panelists. 2. Daniel brought up the Provost's Cyberinfrastructure Survey Report and its significance to our department. It was pointed out that since the Library's Katie Bower designed the survey, it was unfortunate that her name didn't appear more prominently on the report. An important (if not necessarily unexpected) finding of the report is that that the faculty regard "easier electronic access to scholarly information" as the University's most urgent cyberinfrastructure need. Daniel also directed the group's attention to recommendation 2 on page 8: "develop, fund, and implement at least two to three significant projects to connect research tools for searching and access to digital content through Web Service interface standards [... to] enable faculty and students to have seamless access to digital research, instructional, and scholarly material through teaching portals such as YaleInfo and Classes V.2." Daniel pointed out that the provost determines funding, and that new or expanded budget items are a good indication of the University's changing priorities. While this recommendation may be geared more toward ITS than the Library, the emphasis on semantic interoperability, unique persistent identifiers, and integration of content management systems (especially Classes V.2, YaleInfo, and new institutional e-repository), suggests unique opportunities for catalog and metadata librarians. There seemed to be some disagreement among task force members about the extent to which this second recommendation really does involve the library. Daniel suggested that we try to envision ourselves supporting metadata interoperability outside the traditional catalog, since, increasingly, that is where resources are being invested. He also pointed out that while the term "metadata" appears in the report (e.g., on p. 9), the term "cataloging" does not. This has implications for the decision on what to call ourselves, and how best to communicate the work we do within the framework of the university's broader agenda. Daniel brought up the recent LMC/JDC Operational Review Task Force recommendations, the third of which is selectively to "employ 'less than full' or minimal cataloging practices ... [since] not all materials needed to be cataloged to the same level or extent". Matthew pointed out that administrators typically want to limit cataloging standards while reference librarians typically want to raise them. Yale has already tried to reduce cataloging complexity in certain projects, but the prospect of losing access (due to fewer or unregulated access points) is troubling. Meeting adjourned 3pm. |
|||