Bringing Long-Tail Microscopy & Characterisation Data into the - - PowerPoint PPT Presentation

NIF/MA Trusted Data Repository

Projects Datasets Datafiles

* NIF/MA Instrument

User Dataset

• Project ID
• Instrument ID
• Native data
• Implicit/explicit metadata
• Conversions to open format(s)

Handle minting service

Quality Control (QC) Dataset

• QC standard operating

procedure

• QC data

Instrument record

• Instrument

description

• Instrument ID

Data & service discovery portal DOI minting service RAiD service

Published dataset record

• Dataset description
• Dataset ID

Bringing Long-Tail Microscopy & Characterisation Data into the Light

Characterisation: Process of probing and measuring the structures and properties of materials at the micro, nano and atomic scales Long-tail data: Relatively small (KB, MB, GB), unstructured and un-curated Question: What is needed to extend the ARDC-funded NIF Trusted Data Repository solution to include Microscopy Australia instrument data and to facilitate FAIR for both characterisation communities?

* Icons made by Freepik from www.flaticon.com

Key issues

Flexible data model

Projects Datasets Datafiles

Catalogue metadata schemas and vocabularies

• DCC list of Metadata Standards, RDA Metadata

Directory, FAIRsharing.org Standards…

• USID, OME-XML, (EPS) Equipment Data Standard,

Directory Interchange Format, AODN Instrument vocabulary, Core Scientific Metadata Model (CSMD), …

Catalogue file types, metadata extraction tools and file conversion tools

• 60+ instruments recorded to date
• Visits to CMCA + ACMM in November

Data packaging standard

• BagIt
• DataCrate
• RO-Crate

(Research Object Crate)

Standardised protocol for collecting quality data

• V1.0 derived from

NIF Trusted Data Repository Project

Matrix of candidate repository platforms

• MyTardis, Dspace,

CKAN, XNAT, 4Ceed/Clowder, IMS, OMERO, LORIS, …

Community-agreed licenses for data publishing

Findable:

• PIDs and rich metadata for

projects (RAiD), datasets (DOI) and instruments (handle)

• Data & Service Discovery

portal (RDA) Accessible:

• Deposit of quality data into a

trusted data repository service (TruDat@UWA) Interoperable

• Data packaging specification

for interoperability (Data Crate, RO-Crate) Reusable

• Licenses for data publishing
• Open data formats

• Most crucial information must be captured at the

project creation and data collection stages

• Lack of open standards

Ø 100s to 1000s of hours wasted in finding and sharing data, converting between formats, seeking missing parameters and fixing missing values

• Need to support a variety of data repository

platforms

• Agree upon a common data packaging standard to

facilitate interoperability

Ø Metadata schemas and vocabularies Ø Tools for metadata extraction and data transformation

• Cloud-based service for metadata extraction and file

conversion to open formats

• PID services needed: RAiD (Project), DOI (Dataset),

ORCiD (User), Handle (Instrument)

Lessons learnt and findings to date

• 1. Microscopy Australia Data & Informatics Committee 2. National Imaging Facility Informatics Fellow

Project Team:

• Andrew Mehnert1,2 (CMCA, UWA – Project Lead)
• Roger Wepf1 (Director, CMM, UQ; Head, MA D&I Committee)
• Aswin Narayanan2 (CAI, UQ)
• Lisa Yen (Chief Operating Officer, MA, USyd)
• Ryan Sullivan (eResearch Consultant, USyd)
• Matt Foley1 (ACMM, USyd)
• Abby Asomani (Library Information Specialist, UWA)
• Mingfang Wu (ARDC)
• Alexander Joos (CMCA, UWA)
• Instrument managers and technique group leaders at

ACMM, CMM, CAI and CMCA

Acknowledgements