Home

Publishing an Extension Layer

An extension data layer is both a dataset and a work of scholarship. Therefore, the goals for publishing an extension layer include 1) supporting its use and re-use as data (e.g. accessibility, intelligibility, validation, computational tractability), as well as 2) supporting it as a publication within the scholarly communications ecosystem (e.g., discoverability, citability, attribution, provenance)

The CaD@Pitt project does not suggest a single path for publishing extension layers. There are different options available for making data available while still addressing the functions of scholarly communication. This section of the documentation offers general principles and suggestions for publishing extension layers. It also hopes to provide information that will help you evaluate the pros and cons of various repository options.

Preparing your extension layer

The following steps are generalized, and many of them apply to preparing any dataset for publication. Some of the specifics of these steps may be influenced by the choice of publication venue. For example, many repositories will mandate a certain metadata schema, or define a data quality and integrity standard. However, these steps should provide a sense of overall good practice.

• Remove fields copied from base layer, except linking identifier(s)
• Review data for completeness / integrity / legibility
• Create data dictionary / README that describes the individual fields, how they are populated, and any conventions of format
• Create metadata for dataset
  • As mentioned, your publication venue may have its own required schema. However, for an example of a general practice, see the DataCite Schema, a discipline-agnostic dataset metadata schema
  • Because an extension layer has an important relationship to its base layer(s), make a clear reference to those sources in your metadata
  • Remember to include attribution for all contributors to your extension layer

Choosing a venue for publication

There are many options for publishing data, and many for publishing scholarly outputs. Choosing the most appropriate venue for your extension layer will depend on a number of different factors. If you are making this decision for the first time, we offer some pros and cons of three different options: 1) GitHub or similar open code repository; 2) an academic/disciplinary data repository, and 3) your institutional repository (if available).

GitHub or other open code repository

Academic/Disciplinary Data repository

Institutional Repository

---