BLS Case Study
Gary Marchionini
INLS 357

Goal: Improve BLS Web-based Dissemination: 1996-2003
1996-7 Needs assessment (w/Carol Hert, Anita Komlodi)
Interviews (in house, external)
Email content analysis
Transactions and query log analysis
User-task taxonomy
1997-8 UI mockups based on taxonomy, continued needs assessment (Anita Komlodi)
1998-9 Relation Browser prototype and usability study (Ben Brunk, Anita Komlodi)
1999-2000 Relation Browser field test; RAVE extension (Ben Brunk)
2000-01 Replication of needs assessments and extensions to RAVE (Ben Brunk)
2001-03 Evolution of RB database (Ben Brunk, Junliang Zhang)

Lab Stat Help Desk Requests

T-Log analysis (Raw log)

Web requests 1995-2000

Slide 6

Slide 7

Slide 8

Rates of Increase: BLS & LoC

Session Length Oct. 2000

Slide 11

Slide 12

Evolution of BLS website
Minimize clicks to target
Treasures to the surface
User centered vocabulary
25K pages and content management
Academic-Agency partnership model
See Interactions (2003) paper: Digital Government Information Services: The Bureau of Labor Statistics Case. http://ils.unc.edu/~march/interactions_paper03.pdf
See JASIST (2002) paper: Co-evolution of user and organizational interfaces: A longitudinal case study of WWW dissemination of national statistics. http://ils.unc.edu/~march/jasist_BLS_submit.pdf

BLS Jan 1997

BLS Dec 1998

April 99

February 2000

Dec 2000

Nov 2001

September 2003

Information Life Cycle with Feedback Channels

Evolution of Relation Browser
See  http://squash.ils.unc.edu/rave/
See Towards a General Relation Browser: A GUI for Information Architects http://jodi.ecs.soton.ac.uk/Articles/v04/i01/Marchionini/

Slide 23

Relation Browser Version 2.

Relation Browser Version 3

Relation Browser (RAVE) after clicking on Office/Business

Relation Browser (Rave) with results after clicking on Perl

RB+ EIA data, renewables

EIA renewables and usage results

EIA renewables and usage with ‘wind’ text search

Current Work
Continue to generalize RB
Integrate browse and search
Database mappings
Develop automatic techniques for acquiring data
Develop automatic techniques for new ways to slice and dice and provide alternative views