Authority Control IG "Questioning Authorities" pt. I: Tagging and Folksonomies

Questioning Authorities: Adapting Authority Control to the Changing Needs of Library Users. Presented by LITA/ALCTS CCS Authority Control Interest Group. June 25, 1:30 p.m.

Large room, full-ish but not packed at the beginning. As usual with this IG, this is a long, intense program with quite varied presentations, so I will break into two entries.

First three presentations on the agenda deal with

  • user generated organization using vocabularies (tagging and folksonomies, a cool/hot topic);
  • the Library of Congress’s update, given by Beecher Wiggins this time, on the Series Authority Decision and other matters (probably a late add to the program);
  • issues of synchronization of romanized fields with vernacular language encoding in MARC 880 fields when headings are changed by an authority control vendor.
  • These topics show some of the wide range of perspectives that the area of authorities and controlled vocabularies can take – from new insights into psychology, community and language, to ideas for new services or changes and expansion of existing ones, spawned by new technologies, to the mindbendingly complex details of how our current strategies for applying controlled vocabulary get pushed out of whack when changes to rules or features such as parallel encodings are brought into the process – sometimes a topic only an authority control librarian could love.

    Susan Chun from the Metropolitan Museum of Art couldn’t give her presentation on the Art Museum Social Tagging project – her notes will be on the ACIG website a “couple of weeks” after the presentation.

    Louise Spiteri observed limitations of OPACS (one of the themes at this convention) to allow users to customize the organization of saved search results. She then reviewed the landscape of tagging (attaching keywords and other descriptive elements to online bookmarks) and folksonomies (the vocabularies developed collaboratively by communities, using bookmark manager applications) and the study of them she’s undertaking under an OCLC/ALISE grant. Applications like del.icio.us, Furl, and Technorati allow users to contribute tags, re-use the tags provided by others, and search for material using the tags contributed by others.

    Large numbers of people are using these services, and they incorporate new terms more nimbly, include a much wider vocabulary, and reflect the language people use to describe and search for things better than a controlled vocabulary like LC Subject Headings can do.

    If public library catalogs allowed collaborative tagging by users, it could supplement the existing controlled vocabularies, involve people in communities of interest through the catalog, and help librarians create recommended lists and portals. Her studies will include the behavioral aspects of tagging as well as the structure of tags in and comparison with LCSH and NISO thesaurus standards.

    Ross Singer talked about ideas and some plans for plugging the harvested vocabulary from metadata and clusters of terms from tags into a recommender service Georgia Tech is looking into, attached to their OpenURL link resolver. He discussed the features and limitations of some existing tagging applications used with libraries or scholarly literature (CiteULike, Yale Links, PennTags, Connotea, LibraryThing, Amazon Tags). Relatively small academic communities may not be the best basis for building a useful service – if you build it, will they tag? So they are looking at harvesting tags from other places that point to URLs for items in their collections (inhcluding articles) and finding patterns of association with controlled vocabularies from available metadata to populate the service. Sharp points: URLs are weak identifiers. Less used material may also be less likely get tagged.

    Interesting Q/A at the end of the program with Singer and Spiteri responding, about “socially unacceptable language” in tags, catalogers’ traditional intent to be “objective” in applying headings, and the reflection of cultural values in controlled vocabularies (e.g. terms for “same sex union” etc.) – controlled and otherwise. I’m wondering, if complete objectivity seems to be impossible, what should we aim for?