Research Data Services

This page provides an overview of metadata in the specific context of research data and software. There is an excellent guide to Metadata Basics that has been prepared by other librarians at UT Austin, and some of the content is repurposed or copied from that resource.

Metadata is commonly described as "data about data." While easy to remember, this definition is far too vague to be useful. The definitions below provide better explanations in plain English. 

Definition from the National Information Standards Organization (NISO)
"Metadata is structured information that describes, explains, locates, or otherwise makes it easier to retrieve, use, or manage an information resource." 

Definition from Steven Miller, Information and Metadata Lecturer
“Extra baggage associated with any resource that enables a real or potential user to find that resource and to determine value…”

Definition from Karen Coyle, Digital Librarian and Author of Coyle's InFormation
“Metadata is constructed, constructive, and actionable.”

• Constructed - a man-made artifice, not naturally occurring
• Constructive - serving a useful purpose, to solve some problem
• Actionable - can be acted upon, processed by humans and machines

As indicated in the Miller definition above, metadata helps people find resources and determine their value, for whatever need is at hand. This function of metadata is especially critical in digital environments, where humans rely on computer processing for reliable and timely results:

• Metadata facilitates organization, indexing, discovery, access, analysis, and use of print and online resources.
• Metadata enables software agents to navigate and "comprehend" web content.
• Metadata influences search engine results, rankings, and click-through rates.
• Metadata is even more important for non-textual content that isn't readily processed by machines (e.g. images, multimedia, datasets).
• Metadata presence and quality (or the lack thereof) can significantly help or hinder time and money expenditures in research activities.

Researchers will often be familiar with metadata as part of the submission process for manuscripts, datasets, and software. Basic attributes like the title, author information, and keywords are all examples of metadata in research contexts. Metadata are sometimes free-form (e.g., author names) and sometimes selected from controlled vocabularies to ensure consistency and efficient searching within a system (e.g., picking the most relevant topical discipline from a list). Some metadata are common to data regardless of discipline or format (e.g., author information), but many disciplines have subject-specific metadata (e.g., instrument parameters; calibration standards) that should be included as well.