Thing 11: What's my metadata schema?

Overview

Teaching: min
Exercises: min

Questions

• What is metadata and what sort of metadata is critical for research data?

Objectives

• Learn what metadata is.

• Learn why metadata is critical for research data.

• Get hands-on experience with metadata.

Thing 11: What’s my metadata schema? Metadata are the lifeblood for finding and reusing research data. Data is only as valuable as the metadata which describes and connects it. Getting started: What is metadata and what sort of metadata is critical for research data? Learn more: Metadata schema go formal and become standards Challenge me: Hands on with ANZLIC or XML - You choose!

Getting started: Metadata: describing research data

Metadata is your best data friend! Metadata is structured information about a resource that describes characteristics such as content, quality, format, location and contact information. Creating metadata to describe research data is very similar to the process for descriptive cataloguing of library resources. Metadata schema are sets of metadata elements (or fields) for describing a particular type of information resource. Numerous metadata schema exist for describing research data across different disciplines. Probably our most familiar metadata schema are those commonly used in library catalogues and publications repositories such as MARC and Dublin Core.

1. Start by reading the ANDS Introduction to Metadata to understand what metadata is and why is it the lifeblood of research data sharing!
2. Now look closely at this good quality metadata record for research data:

Long-term variation of surface phytoplankton chlorophyll a in the Southern Ocean during 1965-2002 Why do you think this record is considered ‘high quality’? Hint: consider both the type and quality of information provided. What metadata included in this record help discovery and reuse of the data? Look back at the ANDS Introduction to metadata for ideas and think about what we’ve looked at in previous Things such as licensing and sensitive data.

1. If you have time: Sadly, it’s not hard to find examples of low quality metadata describing research data. Read this short 2 page article Avoiding Data Dumpsters - Toward Equitable and Useful Data Sharing on the power of good quality, schema-compliant metadata ( N Engl J Med. 2016 May 11. [Epub ahead of print])

Consider: Why, if metadata is the lifeblood of data discoverability and reuse, is it often neglected or not richly done when data is published.

Learn more: Metadata schema and standards

A metadata standard is a schema that has been formally approved and published, with governance procedures in place to maintain and update the standard. Examples include ANZLIC and DDI (Document, Discover and Interoperate).

Numerous metadata standards exist and the standard chosen to describe resources such as research data should be appropriate to the project or discipline.

1.Start by reading this short guide to Evaluating and Selecting a metadata standard. Keep this Guide open. 2.Choose one disciplinary standard which sparks your interest from the UK Digital Curation Centre’s Directory of Disciplinary Metadata. Use the Guide to critique your chosen standard.

Consider: the particular standard you explored in detail and perhaps add other reasons to recommend (or not) that standard.

Challenge me: Tools for creating metadata

We couldn’t decide on a single activity for Thing 11, so you get to choose between two! Option 1: Hands on with ANZLIC requires you to install open source ANZMet Lite metadata entry tool Is a hands-on experience in creating an ANZLIC compliant metadata record Go to Option 1

tion 2: Hands on with XML does not require you to install software Is a hands-on experience with XML and may help prepare you for Thing 13 (Crosswalks) Go to Option 2

Consider: the metadata creation tools we have explored in Thing 11 - what needs to happen for these tools to have much wider uptake?

Key Points

• First key point.