research objects in scientific publications

Daniel Garijo

Ontology Engineering Group

Universidad Politécnica de Madrid

Madrid, [email protected]

OEG, October 17th 2013

Research Objects in Scientific Publications

http://creativecommons.org/licenses/by-nc-sa/2.0/

Outline

2Research Objects in Scientific Publications

• What is a Research Object?• Why Research Objects?• Research Object model at 5000 ft.• An example from the Life Sciences.• Research Object Overview.

• ORE• Annotations• Extensions

• Eating our own dog food• 6 step example• Cost vs benefit• RO Portal• RO hub

What is a Research Object (RO)?

3

Aggregation of resources that bundles together the contents of a research work


• Process preservation: URIs/DOIs for referencing resources.

• Reusability of any part of the RO

• Repeatability /Reproducibility: redeployment of the method

• Traceability and error detection.

• Attribution: able to cite data and publications of the RO

• Understandability: Links between data, results and annotations.

• Curation: by explicitly exposing the methods of the experiment.

Why Research Objects?

Research Objects in Scientific Publications 4

RO Model at 5000 ft


+ Open Annotation

6

Example (Life sciences)


• Vocabulary for describing Research Objects• Generic• Extensible to multiple domains• Modular

RO Core

OREAO/OA

Research Objects: An Overview


• ORE: Object Reuse and Exchange• Resources can be further specialized according to the domain

The Research Object Model: ORE


The Research Object Model: Annotations


• Additional modules can be described on top of the Core• Evolution• Workflows• Publications• Etc.

RO Core

Evolution

Workflow

Wfprov Wfdesc

OREAO/OA

Other

The RO Model


Eating our own dog food


• Reproducibility is not only an issue

in the Life Sciences.

• In computer science we:

• Run experimental evaluations.

• Reuse previous algorithms

and tools.

• Build pipelines for processing

the inputs and obtain our

results.

• WHY DON’T WE USE RESEARCH OBJECTS?

http://petfoodadvisor.net/wp-content/uploads/2013/02/Cheap-Dog-Food.jpg

How? Minimal step by step example


• When writing a paper about a tool/algorithm/evaluation/survey

1. Create a permanent URL (purl, DOI, etc) for the Research Object

2. Distinguish between the resources that have been used from those that have

been generated.

• A result was generated using/testing an algorithm/tool/framework.

• An ontology was created from the CQs and ORSD of the domain.

• A survey was performed by looking at these papers in the bibliography.

3. Describe the relationships among the resources and their dependencies.

4. Describe the resources in the RO (AO/OA, DC, schema, FOAF)

5. Add all the descriptions to a file (manifest)

6. State that the manifest is describing the purl of the RO.

7. You are done.

Annotating RDF/RDF-a can be exhausting


http://i3.kym-cdn.com/entries/icons/original/000/006/725/desk%20flip.jpg

• Cost vs benefit.

• It takes time, BUT

• All your resources become easily

referenceable:

• You provide context to your research

• You ease the task of reusing your

software/papers

• There are tools to ease the task.

• You don’t need to annotate everything!

• The more you annotate, the better

your results will be understood.

• Creation of ROs• Annotation and visualization of ROs

http://sandbox.wf4ever-project.org/portal/home


• Many approaches exist for making your stuff persistent• Figshare, Purl, Datacite, RO Portal, etc.

• Rohub.linkeddata.es is a page (under development) for pointing at the resources created in ROs of the OEG.• 3 ROs at the moment.• Content negotiation (TTL and more)• Annotated in RDF-a

• ROs as web pages with purls!• http://rohub.linkeddata.es/ro-svmworkflow/

(http://purl.org/net/svm-opt-research-object)

rohub.linkeddata.es


http://esp.habitants.org/campana_cero_desalojos/jornadas_mundiales_cero_desalojos_-_por_el_derecho_al_habitat_2013/work_in_progress

RO-ify your work!http://researchobject.org/

Join the discussion…W3C Research Object for Scholarly Communication Community Grouphttp://www.w3.org/community/rosc/

Acknowledgements: Wf4Ever project, Idafen Santana

Daniel Garijo

Ontology Engineering Group

Universidad Politécnica de Madrid

Madrid, [email protected]

OEG, October 17th 2013


http://creativecommons.org/licenses/by-nc-sa/2.0/

research objects in scientific publications

Education