Innovative Friends: Alex Wright

I’m psyched that Alex Wright now calls New York home. He is one of the few thinkers of technology and information that can simultaneously ply his trade during the day and theorize with the best in his spare time. I’m looking forward to his new book Glut: Mastering Information Through the Ages in which he surveys the history of methods for managing oceans of information for clues of where we’re headed.

And I’m tickled that he received a review from James Burke of BBC’s Connections, how appropriate…

This is a must-read for anybody who wants to understand where we’ve been and where we’re going. A lucid, exciting book full of flashes of surprise about how we’ve done it all before: prehistoric beads as networking aids, 3rd century random access systems, 7th century Irish monastic bloggers, 11th century multimedia, 16th century hypertext. I wish I’d written it!

Notes from ASIS&T Annual 2002

Notes from ASIS&T Annual 2002 Monday, November 18
Just spurious notes from what I saw, in general a very good conference:

Plenary with Lee Strickland Thomas Blanton: Openness and National Security
I went in thinking this talk would be boring, but found it very interesting and by the end I was convinced we all need to spend at least a bit of time thinking about how to handle the openness of information within our organizations. I found Tom Blanton particularly insightful, pulling together observations from various inside Washington venues to paint a picture for outsiders.

U.S. government has increased powers to seize information following 9/11. If they show up, we should know

  • What to do
  • Who (lawyer) to call
  • How to handle the event

Law enforcement can intimidate with badges and guns, but of course they need the appropriate court order to seize anything.

Session on Metadata
Elizabeth Liddy and her colleagues at Syracuse are doing great work with metadata. She covered three projects.

  • Automatic metadata generation through natural language processing.
    • A key factor in improving results was informing the system with domain-specific information
    • Used Dublin Core and GEM
    • Information Extraction method

    It’s sponsored by NSF funding spread across partnering universities.

  • Standard Connection used a standard compendium that maps together various state school standards. The system then attempts to automatically map a resource to standards.
  • Metatest is a two-year study questioning our assumptions about the value of metadata. One section of it was interesting in that they combined eye-tracking studies and think-aloud protocols.

Terrance Smith presented on knowledge representation of scientific info. He used three projectors in class to display three perspectives on info: outline, metadata, and data. The layout of the information suffered, exposing the difficulty in taking highly structured concepts and relationships and presenting them naturally. He observed that concepts are harder to index than terms

Marcia Lei Zeng of Kent State: There’s a trend toward enhancing standard metadata formats like Dublin Core with domain-specific fields. This parallels development of custom markup languages and metadata formats. The idea reinforced the importance of semantic markup. Most of the time I’m dealing with chunked content, each chuck having its own metadata. But if forced to work on the document level it could be important to sync your metadata format and markup language.

Eric Miller and James Hendler of the W3C presented an overview of the semantic web, RDF, and OWL. Nothing new, but great to hear it being evangelized at ASIST by two smart, effective presenters. SWAD (semantic web advanced development). Check out their initiatives. He repeated the formal ontological features that differentiate them from controlled vocabularies, like semantic restrictions on property relations, range, domain, cardinatlity, logical sets, inverse relationships (same but from the opposite view, like parent of and child of) etc. Some German organizations who signed up: DFKI, Forschungszentrum Infomatik. The layer cake continues to be an important model, they both used it in their presentations – We need to acknowledge in these activities that creating metadata (any kind) is hard and it needs to be reused enough to provide value or else it’s not worth creating – Miller is using the Razor collaborative spam app

Also see U of Maryland’s Mindswap – The Semantic Web Research Group

IA Case Studies I’d Like to See

1. Slashdot. For the interplay of how information feeds into the interface design, how the whole system is tweaked to serve very high traffic efficiently, and the interplay of those two challenges.

2., for the brilliant-yet-evil way they suck you in, offering you a taste, getting a little info about you, giving you a little more, getting a little money from you, and so on. The lure is very strong; I can understand how they afford all those pervasive banner ads. For students of permission marketing this is a must experience.