Data discovery with DATS: exemplar adoptions and lessons learned

Area of Science:

• Data Science
• Bioinformatics
• Information Science

Background:

• The Data Tag Suite (DATS) is a data model designed to standardize dataset description, indexing, and discovery.
• It utilizes schema.org for web discoverability and powers DataMed, a prototype data discovery index.
• Previous experience with DataMed involved indexing over 60 diverse repositories.

Purpose of the Study:

• To evaluate the fitness of the DATS model for describing diverse datasets.
• To refine DATS based on practical application and user feedback.
• To identify best practices and future development paths for DATS.

Main Methods:

• Mapping three additional diverse data sources to the DATS model.
• Involving data representatives and experts in the mapping process.
• Gathering feedback from users and implementers of DATS.

Main Results:

• DATS demonstrated fitness for describing new, diverse data sources.
• Implementation guidelines and best practices for DATS were developed.
• Specific needs for dataset indexing, particularly for clinical and observational data, were identified.

Conclusions:

• The DATS model is effective for enhancing dataset discoverability.
• Practical application and feedback are crucial for model evolution.
• Further optimization is needed to address complex data types like clinical information.