WAND Taxonomies provide a foundation set of relevant terms businesses can use to populate their business glossary in IBM Cloud Private for Data (ICP4Data).
ICP for Data is an platform designed for data science, AI, and machine learning.
An important element of ICP for Data is the ability to label data with a consistent terminology that has clear meaning for a business. This ensures that the entire business is clear on the meaning of a term and is labeling data correctly, avoiding errors and misunderstandings when performing data analysis and creating AI models.
Consistent terminology for labeling is defined in a business glossary. This is where the WAND Taxonomies come in. WAND Taxonomies contain an extensive set of terms for each industry vertical and business functional area covered. Instead of creating a business glossary from scratch, ICP for Data users can import a WAND Taxonomy and then edit and customize the concepts.
Once a business glossary is defined, ICP for Data can be used to catalog and label data with the glossary terms.
In ICP for Data, taxonomies are called Data Dictionaries or Business Glossaries. Data Dictionaries can be imported via CSV or via XML. XML provides an advantage over CSV in that the XML format allows a user to select override settings which can assist when importing updates to a vocabulary.
The Business Glossary function can be accessed underneath the "Organize" section of the menu.
The WAND Taxonomy Library Portal has both the CSV and XML export formats available so WAND Taxonomies import seamlessly into ICP for Data.
Once a business glossary is created, editing of the glossary can take place. This may mean adding new terms, creating or editing definitions, removing terms, or creating relationships between terms.
Third Party Resources
|I don't see the "Organize" section in the menu. What am I doing wrong?
Make sure that the Data Catalog module has been installed in your instance of ICP for Data.
To learn more, contact us online or via email and we'll respond within 24-48 business hours.