Skip to main content
digdir emblemdata.norge.no
Find dataShare data

1. Data mapping

Note: The organization you work for may follow the Order in Your Own House guide, and you are part of this process. This is a thorough and useful guide for getting an overview of the data your organization holds and can share with others. Even if your organization hasn't completed all the steps in the guide, you can still publish information about the datasets you've already mapped. The descriptions can gradually be improved and expanded as you progress in your work with Order in Your Own House.

First, you need to know which dataset you want to describe, and information that might be useful to have ready for each dataset includes:

  • The title of the dataset.
  • How you want to describe the dataset textually.
  • Translations of the title and description so that they are available in Bokmål, Nynorsk, and English.
  • Which organization owns and publishes the dataset, and the organization's registration number.
  • Contact information (email) for those responsible for the dataset.
  • Whether the dataset should be offered openly,
    • and if so, whether it is available for file download or via an API.

Tip: Questions regarding whether the dataset should be open or closed and which licenses should be used are often something legal experts must decide. While these clarifications are being made, you can still publish all other information. Once you have agreed on whether the dataset should be open, how it should be made available, and which licenses apply, you can expand the description with this information.

Example of data mapping: AI projects in the public sector

Digdir offers a dataset that provides an overview of AI projects in the public sector. In the description of this dataset, we want to include at least the following information: - The title is "Artificial Intelligence - Overview of Projects in the Public Sector." - The Bokmål description is "An overview of artificial intelligence projects in the public sector. The overview is not complete." - The translations to Nynorsk and English will be added later. - The publisher is the Norwegian Digitalisation Agency, with registration number 991825827. - Those responsible for the dataset at Digdir can be contacted via the email postmottak@digdir.no. - The dataset can be linked to the topic "public administration," using a code from a shared code list. - How this is done will be shown later in the guide.

This is all the information needed to cover the mandatory fields for a dataset description. Additionally, there are a few other things we would like to include to make the dataset easier to find and use for others:

  • The dataset has its own webpage: https://data.norge.no/kunstig-intelligens
  • The dataset is available for download from Digdir's Github page.
  • The dataset is offered/distributed as a CSV file.
  • The dataset is open and available, and licensed under the CC0 "No Rights Reserved" license.
  • The content of the dataset is in Bokmål.
  • The dataset was published on February 23, 2023.
  • The dataset can be associated with the keywords "artificial intelligence" and "public sector."
  • The dataset is based on information collected from third parties.
Tip: Using shared codes and keywords makes the dataset easier to find in searches.