Centrelink Thesaurus - a case study - Kathleen Lazzari Manager, Resource Discovery Centrelink Abbreviated version of a paper given at the Information Architecture.

Centrelink Thesaurus - a case study -

Kathleen LazzariManager, Resource Discovery

Centrelink

Abbreviated version of a paper given at the Information Architecture & Taxonomies Conference 6-7 November 2003, Sydney

actKM Meeting 3 February 2004

2

Topics• Definitions

• Centrelink – the organisation and I&T context

• Centrelink Thesaurus– The taxonomy : content and technology– Steps in development– Search, navigation and other benefits– Current enhancements– Future opportunities

• Training and promotion

3

Definitions• Controlled vocabulary

– Managed set/list of words or phrases

• Taxonomy– Classification tool, pre-defined system– Defines groups and subgroups– Framework for discussion, analysis, information retrieval– Provides terms of a domain – in context– Represents a structured relationship

4

Definitions• Thesaurus

– Reference tool which groups synonyms, defines relationships

– Controlled vocabulary– Taxonomy– Examples: www.lub.lu.se/metadata/subject-

help.html

5

Centrelink - the organisation

• Australia’s largest supplier of Commonwealth govt products, services

• Australian govt’s one stop shop for services• I &T Network – World’s ninth largest• 27 000 staff, more than 450 sites• Australia’s largest single purpose call centre

network• Client departments – partners

6

Centrelink – the organisation (cont.)

• Facts and figures:– www.centrelink.gov.au– 6.4 million customers– 12 million online transactions daily– Per annum:

• $53.4 billion paid• Almost 100 million letters to customers• 5.3 million new claims• Over 23 million phone calls

7

Centrelink – the organisation (cont.)

• ‘Government and community sensitive’, often rapid change– Policy, legislation, budget initiatives etc– Ansett collapse, Bali bombings, drought, ACT

fires• Need to ‘get it right’• Major assets: people, information

8

Centrelink – I&T context

• Constant, often rapid, change

• Substantial costs to change tools, platforms

• Lotus investment– Desktop tools– Internet platform, search engine, approx 2000+ pages

• Intranet– Over 45 000 pages– Unix, flat HTML, Cold Fusion applications, Verity search

engine

9

Centrelink – I&T context

• Business intelligence – growing importance• Government injection of funds for I&T refresh• Web content management system implementation

– Sun/divine’s content server

• Distributed authors, publishers• Dissemination of / reliance on data and

information• Exhaustive, specific info retrieval required

10

Centrelink – I&T context (cont.)

• AGLS metadata standard : good practice to policy

• Content and records management

• Record keeping metadata standard

• Organisation taxonomy – Centrelink Thesaurus

11

Centrelink Thesaurus:the taxonomy

• Large, hierarchical list of terms and names– groups, subgroups, synonyms– thirty ‘top’ terms– to approx 3 levels

• Defines relationships– broad, narrow, related

• Defines preferred terms (approx 3000)– maps ‘non-preferred’ terms, abbreviations,

acronyms

12

Centrelink Thesaurus:the taxonomy (cont.)

• Tracks changes over time– former name, later name– scope and history

• Centrelink business coverage– subject focus– includes payments, teams, programs, govt.

bodies, organisations

13

Centrelink Thesaurus:the technology

• DB2 database

• Corporate standards and processes

• From Cold Fusion to API with Java (J2EE)

• Constructed in accordance with ISO 2788

14

Steps in development

1. Identification of need, benefits

2. Prioritisation of objectives

15

Steps in development (cont.)

3. Environmental scan– standards and manuals

• ISO 2788 • Aitchison, J. Thesaurus construction and use.

– existing taxonomies– online information– stakeholders

4. Acquisition of expertise

5. Identification, specification of tools, database, methodology

16

Steps in development (cont.)

6. Population of taxonomy

• base taxonomy (APAIS) : National Library of Australia

7. Documentation of processes, procedures, policy

8. Initial launch and implementation (July 2001)

9. Database redesign (API, Java) 2002-03

10. Planned re-launch late 2003

17

Search, navigation and other benefits

Discovery of resources in web environments Support of indexing for information retrieval

• controlled values for metadata

Searching: high recall, precision, more options, consistency

Navigation: enhanced, consistent• index generation• browse

18

Search, navigation and other benefits

Examples of what can be managed• Youth / Teenagers / Adolescents / Young

people• Aurally disabled / Deaf• Website / Web site• ATO / Australian Taxation Office/ Tax

Office• DFaCS? / FaCS? / DFCS?

19

Search, navigation and other benefits

New language created/adopted

Language standardised in Thesaurus

Thesaurus useda language authority

Corporate language appears in content

Content indexed with Thesaurus terms

Easier, consistent search, navigation

and retrieval

20

Other benefits

Information management• efficiencies in content management

Support internal and external sites• intranet, staff portal development • cross agency consistency, compatibility• online services & information for customers,

clients• eg: fed.gov.au, customer themed portals

Support govt initiatives, eg Government Online Strategy

21

Current enhancements Database tables redesigned to enable

– taxonomy subsets• create once, use many

– efficient management of vocabularies

– accommodation of definitions (glossary function)

– integration with web content management system (authoring, indexing, harvesting)

– integration with other applications

22

Future opportunities

• Browse and search tool– big picture of enterprise information

• Facilitate end user 'always right'

• Positioning for advances in 'auto categorisation' software

• Increasingly consistent, correct language

• Integration with ‘structured’ info tools, products, search, navigation

• Supply of business definitions

23

Other info stores

Future opportunities (cont.)

FunctionThesaurus

Subject Thesaurus

Data dictionary

End user search/browse/view

24

Training and promotion

• Staff who maintain the taxonomy mustbe appropriately qualified and trainedbe advanced in the language of the taxonomyhave advanced communication skills

• Authors, publishers will require training

25

Training and promotion (cont.)

• Management needs to understand commitment, opportunities, efficiencies

• Particularly need to educate, promote with– content managers– records managers– knowledge managers– CIO– I&T staff managing above tools

26

Thank you

Kathleen Lazzari

Manager, Resource Discovery

Data Shop, Centrelink

Ph 02 6244 1438

© Commonwealth of Australia