An ORCID based synchronization framework for a national CRIS ecosystem

An ORCID based synchronization framework for

a national CRIS ecosystemJoão Mendes Moreira, Alcino Cunha, Nuno Macedo

Our problem

• Research outcomes scattered through various services and data formats

• Heavy burden on researchers and managers

• Inconsistent information and impact assessment

2

Our goal

• Input once, reuse often!

• Automatic synchronization between services

3

PTCRIS in context

4

PTCRIS in context

CV

4

PTCRIS in context

CV

OA aggregator

4

PTCRIS in context

CV

OA aggregatorDSpace

4

PTCRIS in context

CV

OA aggregatorDSpace

OJS

4

PTCRIS in context

CV

OA aggregatorDSpace

OJS

4

PTCRIS in context

ORCID

CV

OA aggregatorDSpace

OJS

4

PTCRIS in context

ORCID

CVResearcher

ID

OA aggregatorDSpace

OJS

4

PTCRIS in context

ORCID

CVResearcher

ID

OA aggregatorDSpace

OJS

Scopus

4

PTCRIS in context

ORCID

CVResearcher

ID

OA aggregatorDSpace

OJS

Scopus

4

A non solution

ORCID

CVResearcher

ID

OA aggregator

DSpace

OJSScopus

5

A non solution

ORCID

CVResearcher

ID

OA aggregator

DSpace

OJSScopus

5

A non solution

• Many synchronization procedures per service

• Implementation and maintainability nightmare

• No clear flow to achieve global consistency

6

Our solution

ORCID

CVResearcher

ID

OA aggregator

DSpace

OJSScopus

7

Our solution

ORCID

CVResearcher

ID

OA aggregator

DSpace

OJSScopus

7

Our solution

• One synchronization procedure per service

• Easier to implement and maintain

• Clear flow to achieve global consistency

8

Why ORCID?

• High coverage of predefined requirements

• High interoperability with external sources

• High coverage of the national research community

• Sustainable and open source

9

Modular synchronization

ORCID

10

Modular synchronization

ORCID

10

Modular synchronization

• IMPORT: harvest research outputs from ORCID

• EXPORT: feed research outputs to ORCID

ORCID

10

Main challenges• Mismatch between data models

• ORCID may contain several versions of the same output

• They are grouped together based on UIDs

• (Fast evolving) API still does not reflect this data model

• Inconsistent, sometimes incorrect, meta-data

• (Still) no clear mechanism to determine the user preferred version

11

Design methodology

• Focus on the what before the how

• What is the desired notion of consistency between each service and ORCID?

• How can a synchronization procedure be implemented to enforce such consistency?

12

Design methodology• Use formal analysis methods and tools

• To verify desirable properties of all artifacts

• To generate scenarios for validation and testing

• “The first principle is that you must not fool yourself, and you are the easiest person to fool” Richard Feynman

13

Scenarios

Article1 doi0 handle0

Paper1 doi1 eid1 handle1

Work1 USERdoi0

Work2 SCOPUS

doi1 eid1

Work3 USER

eid1 handle1

14

Scenarios

Article1 doi0 handle0

Work1 USERdoi0

Work2 SCOPUS

doi1 eid1

Work3 USER

eid1 handle1

Paper1 doi1 eid1 handle1

doi0

14

Scenarios

Article1 doi0 handle0

Work1 USERdoi0

Work2 SCOPUS

doi1 eid1

Work3 USER

eid1 handle1

Paper1 doi1 eid1 handle1

doi0

Work4 PTCRIS

doi1 eid1 handle1 doi0

14

Scenarios

Work1 USERdoi0

Work2 SCOPUS

doi1 eid1

Work3 USER

eid1 handle1

Paper1 doi1 eid1 handle1

doi0

Work4 PTCRIS

doi1 eid1 handle1 doi0

Article1 doi0 handle0

doi1 eid1 handle1

14

Prototype

• Demoed at the FCCN community annual meeting in February 2015

• Show compelling and real use cases

• Four PTCRIS services involved: CV, OA aggregator, DSpace, and OJS

15

Prototype

16

What’s next?• We expect to have certified implementations of this

specification in DeGóis (CV), RCAAP (OA aggregator), SARI (DSpace), and two local CRIS systems by the end of 2015

• Upgrade the specification to ORCID API 2.0

• Provide synchronization as a service?

• Automatic (model-based) testing from scenarios?

• H2020 proposal around this topic?

17