Interoperability: recombinant potential Lorcan Dempsey ‘opportunities for applied research on the creation, management, preservation, and use of digital.

Interoperability:recombinant potential

Lorcan Dempsey

‘opportunities for applied research on the creation, management, preservation, and use of digital

content’

IMLS workshop

17/18 March 2003

Overview

• MARC

• MARC-XML, MODS, Dublin Core, Onix, LOM

• EAD, TEI, DC, MARC

• METS, SCORM

• DDI, FGDC, ..

• MARC AMC, EAD, DC, RSLP

• OAIS, METS, OCLC/RLG, …

• Z39.50, SRU/W, Xquery, …

• SOAP, WSDL, UDDI, …

• GIF, TIFF, PNG, JPEG, …

• XML, RDF, ..

• DDC, LCSH, LCC, TGN, AAT, …

• PURL, DOI, ISTC, URN, …

• XRML,ODRL,..

Overview

• Remarks on interoperability

• Institutional perspective

• Interoperability frameworks

• Concluding remarks

• Inconclusive!!!

– Different mental models

– The words of things entangle and confuse

• document, repository, archive, digital library, registry, …

“Creative knowledge you can put in your pocket”

Interoperability as recombinant potential• Can I …

– … add a document to a repository?

– … add a repository to a distributed query?

– … fuse metadata from one repository with another?

– … assemble these resources into a learning package?

– … embed an interactive service in my exhibition, my reading list, my campus portal?

– … ingest a content package into an archive?

– … take a content package out of an archive in 10 years time?

– … navigate several databases by subject, by name, by place, by resource type, by educational level?

– … cite a document in a repository?

– … bring resources into my own workspace?

• With …

– … as little custom work as possible

– … as little precoordinated agreement as possible

Licensed collections

Aggregations

Virtual reference

user environmentsresource environment

Library serviceLibrary serviceenvironmentenvironmentLibrary serviceLibrary serviceenvironmentenvironment

Institutional repository

Digital collections

E-reserveCatalog

lab books

exhibitions

PDAs

learning management systems

campus portal

course materialtext book

new scholarly resources

readinglists

Different perspectives

• Disaggregating scholarly publishing

– Linking, Identifiers

• Libraries, archives and museums digitizing unique collections

– Content and metadata models

• Research and learning materials

– Eprints

– Research data

– Learning materials

– E-portfolios

• Service model: multiple environments of use

– Exhibition, workbench/lab book, learning management system, portal, interpretive essay, …

• Metadata models

– Syntax

– Semantics

– Values

• Content packaging models

– METS, SCORM

• Collection model

– Heirarchical, multiple items

– Provenance, context

– Interpretive, organized knowledge, evidential

• Use model

– Tools required, …

• Resources (Objects and collections)

– “content”

– “metadata”

– Rights

– People

– Institutions

– Concepts

– Places

– Resource levels

– Educational standards

– …

• Services

– Query

– Disclose

– Get/put

– Validate

– Ingest

– Annotate

– Request

– Register

– Migrate

– …

• Interoperability Infrastructure

– Registry• Publish structures/schema

– Directory• Services, collections

– Resolution• Relate other resources

A changed world

• A pervasive network environment requires a more light-weight, fine-grained approach to combining and recombining resources and services.

• “Web services”

• See pre-workshop comments.

• We have done most work at metadata level (and access to metadata), but even there …

Some metadata examples

• Equivalence?

– Network digital library of theses and dissertations

• Management of identity and difference

– E-prints UK

• Consistency

– Schema transformation

– Harvesting LOM

• Knowledge organization

– Articulate resources and users by subject, place,

• Major issues

– Different communities of practice (case law)

– Leverage accumulated investment in knowledge organization technologies or see them wither

– Harvesting nascent – need agreements

– What can be automated

– Precoordinated vs dynamic recombination

An institutional perspective

• High acronymic density a source of bewilderment

• Makers and takers

• State of the art keeps receding over the horizon

• Balkanization leads to sub-optimal approaches

Rich

Descrip

tion

Sim

ple

Desc

rip

tionItemCollections

Dublin Core

RSLP

OAI set record

TEI

VRA Core

ONIX MARC 8

CSDGM

Institutional issues

• Organizational domains

– Library

– Archive

– Press

– Learning management

– Departments

– Computing

• Interest domains

– Grid/Internet2

– Digital library

– Learning management

– Disciplinary focuses

• Shared services

– A variety of repositories

– Identity management

• Institutional investment

– What confidence?

– Vertical silo within institution -- wheels within wheels

– How do you manage investment across silos

– Current metadata creation and content management practices unsustainable.

Interoperability frameworks

• Reference model

• Functional architecture

DirectoriesVocabularyCompetencyMetadata

Repositories

Organizations Traders

Acc

ess

Man

agem

ent

MA

NA

GE

RIG

HT

S O

BL

IGA

TIO

NS

CO

NT

RO

L A

CC

ESS

AU

TH

EN

TIC

AT

E AU

TH

OR

ISE

AU

DIT

Pro

curem

ent

NE

GO

TIA

TE

TR

AD

EM

AK

E P

AY

ME

NT

SEARCH

Learner Creator Infoseeker

Repositories

AssetsMetadata

Resource Utilizers

DISCOVER

REQUEST

USE Presentation

Mediation

Provision

People

Agent

RE

SOL

VE

Registries

STORE

STORE EXPOSEMANAGE STORE EXPOSEMANAGE

DELIVER

(Query, Browse, Follow Path)ACCESS

GATHER

PUBLISH

MANAGE

ALERT

EXPOSE

IMS digital repository interoperability specification

JISC Information Environment architecture

Interoperability frameworks – the big IF

• Positive

– Hedging our bets

– Common frame of reference

– Articulate business, technical and service discussion

– Orient and motivate development

– Suggest roles and responsibilities

– Examples• OAIS

• OKI

• IMS

• JISC IE

• Portal envy

• Questions

– At what level of granularity

– Cross-domain?

– Unhelpful circumscription?

– Stifles thought

Some concluding thoughts• Institutions and incentives !!!!!

– Our apparatus for managing standards is dysfunctional.

– What incentives and support for creating recombinant schema, resources (metadata and content), and services would be useful?

• Interoperability plus time = preservation.

• Service architecture

– We need a greater focus on exposing and combining services in user or system environments.

– What level of explicit “interoperability framework” is useful?

– What interoperability infrastructure is useful and how is it sustained (registry and directory)?

• Make data work

– We need better approaches for managing resource identity and difference.

– Mining – needs data

– Can we make knowledge organization technologies more effective in a network environment?

• Why has digital library research had so little impact in operational environments?

Thank you!

Lorcan Dempsey