Frequently Used Pages

Emerging Technologies Group (ETG) Meeting

notes taken by Starr Hoffman

6.20.06

minutes are on shared drive:

H:\SHARED\STAFF\Committees and Work Groups\Emerging Technologies Group

ETG doesn't perform projects
ETG they identifies projects, goes over issues, makes recommendations, and helps with the group performing the project
need Charter Scope Statement
- have six months (until October)
- will discuss again in August

discussion about moving them online (whether in a blog or other format)
comments from library employees on the forum
could use wiki for this discussion, but prefer forum (all info in one place, hosted by library)
might require a login, similar to some InHouse articles
may be blog format, but should easily incorporate RSS regardless
possible project managers: Diane Wahl, Gay Woods (very interested in blogs)
these ideas are not yet written down; should add these to the forum
not quite ready to hand this project off

in general, university-wide (like MyUNT) portals are more in favor than library portals
however, in MyUNT, we'd probably we at the bottom of the far right column (not very prominent)
at best, we'll probably get a link out to the mylibrary portal
what services should be offered, and are their other methods of offering them?
login through LDAP may only provide declared major (not always current) and contact information
would like to receive info on student status (year) and courses for that semester (could then display relevant reserves holdings)
optionally, users could fill in blanks: major, status--use these to set up preferences
can use mylibrary to keep track of internet searches & sources
suggest ER databases based on past history & course info
mymillennium doesn't seem as flexible as mylibrary, but could perhaps be combined with it

Upcoming TechTalk on the topic (8.22)

this system would work aside Innovative (ILS)
trying for a more intuitive system for students
trying to head toward single search box, more like Google (only Endeca has data dumped into it, while Google harvests data)
still no database-level searching; suggests db's based on topic
similar to "related" search concept

Endeca
- example: NCSU catalog (info)
- very expensive (Endeca $800,000 without consulting fees)
- UT moving to ILS instead of home-grown
- "guided browsing" or "browse navigation/searching"
Siderean "faceted searching" (RDF / RSS)
several vendors are courting Cathy; all msgs are fwdd to Mark
Berkely "open source-ish" model
- Mark has tested this with a collection of 30,000 photos organized by various facets (format, year, people, keywords)
- small tests of MARC records in this model reveal cataloging problems
big vendor solutions aren't very good; SFX is okay
LOT model (cheap through Keystone software)
what are Amazon / Barnes & Noble using for their search?
- B&N uses Endeca

each week, dump all catalog info (and ER, and digital collection) into this system
would include all libraries content:
- catalog
- digital collections
- ER databases (won't search individual articles, just database by topic)
not truly a federated search (in which use protocols like Z39.50 to retrieve info from various databases)
federated search: LOT performs realtime queries & points to data in different locations
this model is problematic to implement in a system like ours
facted search: Endeca searches (one location) within info that is dumped into it once a week (from all library content)
facets would be based on MARC records (subject, publisher, etc.)

iii will make getting core dump of MARC records difficult
could simply add content to catalog, except that you'd only get a metadata search of digital collections (lose full-text keyword searchability)
what about Webfeat?
- Mark will explore this
Mark will look into this more, gather info at ALA, and get literature on it

Back to the Front Page