Tutorial: Using Dublin Core Tabular Application Profiles (DCTAP) (Open Session)
- Starts at
06 Oct 22 21:00 UTC
- Finishes at
07 Oct 22 00:00 UTC
- Virtual Conference Room A
- Sam Oh
An application profile defines metadata usage for a specific application or usage. Profiles are often created as texts that are intended for a human audience and these texts generally employ tables to list the elements of the profile and related rules for metadata creation and validation. There has not been, however, to date a machine-actionable format for profiles. The DCTAP combines the need for human-readability and machine-actionability: it is a simple table format (with underlying comma separated values) with 12 elements that provide a core description of an application profile. The TAP is being developed as a Dublin Core working group. Documentation and examples can be found at https://github.com/dcmi/dctap.
Target audience: All levels
Tutorial style: Demonstration with hands-on practice and software experience
After a brief presentation of the DCTAP structure, participants will create a simple DCTAP using any spreadsheet program. A few TAPs will be available or participants can bring their own metadata to model. During this exercise we will also describe TAP's extension possibilities. Finally, participants will use an available Python program to convert their TAP to one of these formats: JSON, YAML, or text.
Any prior knowledge required (eg, RDF, programming languages)
General knowledge of metadata. Familiar with any spreadsheet program (including Google Sheets). Basic comfort with using the command line to install and run programs.
Whether participants must (or should) bring laptops or install software beforehand.
For the hands-on portion participants should have a laptop with a spreadsheet program installed, if preferred, or access to an online spreadsheet service like Google Sheets. If they wish to model their own metadata they should have that documentation available. Some of the exercises will make use of a Python program that requires Python 3.7 or higher. The installation instructions for the Python program can be found at https://dctap-python.readthedocs.io/en/latest/ and participants are incouraged to install the program prior to the tutorial. The program will be demonstrated at the tutorial and running the software on ones own laptop is recommended but optional.
Sam Oh is the current executive director of DCMI, professor emeritus of Sungkyunkwan University Korea, affiliate professor at University of Washington, Seattle USA. His expertise includes data modeling, metadata and ontology design. He has consulted many companies and government sectors in Korea. He served as a past chair of iSchools and now is the iSchool ambassador. He is the chair of TC46/SC4 (Technical Interoperability), also chaired both TC46/SC9 (Identification & Description) for 6 years and ISO/IEC JTC1 SC34 (Document Description and Processing Languages) for 9 years.
Karen Coyle is a librarian with decades of experience with metadata. Karen has been active in the development and management of metadata standards serving a variety of communities. She is currently investigating the possibilities offered by the semantic web and linked data technology, working with the Dublin Core Metadata Initiative and taking part in standards efforts of the World Wide Web Consortium.
University of Alberta Library
John Huck is a metadata librarian at the University of Alberta. He is the chair of the Dataverse North Expert Group within the Digital Research Alliance of Canada's Research Data Management Network of Experts. His areas of experience include geospatial metadata, research data metadata, and semantic web standards and technologies.