Skip to Main Content

Metadata Creation

Brief guide that explores the topic of metadata: What are metadata? What kinds of metadata are there? What is the process of creating metadata?

Step One

When preparing to describe your resources, there are a number of questions that you will want to consider:

  • What you are describing?

Are you describing a physical object, a digital object, a digital representation of a physical object?

  • What kind of information do you want to record?

Who is my audience? What information is needed to identify the resource? What information is needed to properly contextualize it? How do I want people to find it or interact with it? How do I expect them to search for or discover it? How do I expect to use it? How do I expect others to use it now and in the future? What information is required to communicate who owns it, who can use it and to what extent?

Step Two

As you begin to answer the questions presented in Step One, list out the information that you would like to include as data points, e.g., title, subject, access rights, etc. For example, if you are wanting to overlay images onto a map, you will want to record coordinate data. This is your metadata wish list.

Step Three

Consider the descriptive information or metadata that you may already have: Which elements or what kind of information are recorded or represented there? Is information missing about your resources? Is there information that would challenging to find or create?

Step Four

Find your "golden minimum."  Determine what information is essential to facilitate discovery, identification, and to give sufficient context, but no more. What exactly is the golden minimum in the space of your project depends on your project goals and available resources.

Step Five

Finalize your list of data points. Choose to codify this list as your own metadata schema or map it to an existing schema, such as Dublin Core.

Step Six

Decide whether you want to make use of data value standards (controlled vocabularies, thesauri, encoding or formatting standards). If so, which standards would apply to which fields? Alternatively, you can create your own data value standards, such as, a subject vocabulary specific to your topic or collection of resources, or a controlled list of names. Document your decisions as your best practices.