liveBook · Manning

Data discovery ticket

As explained in Delivering Machine Learning Projects: From design to deployment

A data discovery ticket is a project work item for collecting and documenting information about data resources and their constraints.

A data discovery ticket turns early data investigation into a concrete project work item. It reduces delivery uncertainty by revealing what data exists, what it contains, and what using it will require. section 2.4

Create the ticket around a representative data sample and document:

  • Statistical properties, such as the data’s measurable characteristics. section 2.4
  • Nonfunctional properties, including scale, speed, and historical coverage. section 2.4
  • System properties, including location, infrastructure, and operational purpose. section 2.4

Use open-ended questions to find overlooked sources, validate known datasets, identify deficiencies, and estimate cleaning, supplementation, or limited-data work. A full dataset may be impractical initially, so begin with a sample. section 2.4

How Delivering Machine Learning Projects: From design to deployment covers it

How deep it goes
Specifies a sample-based documentation checklist and stops short of full-data extraction or comprehensive validation.
Examples
Uses a data sample and a PS6 discovery ticket containing statistical, nonfunctional, and system properties.
What it assumes
Assumes access to client data knowledge and a usable sample while teaching what evidence to collect from both.
What you can do after
Evaluate data readiness by documenting sample characteristics, infrastructure constraints, gaps, and likely preparation work.
How it connects
Connects data discovery to data engineering, data quality, data sources, and machine learning solutions.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Delivering Machine Learning Projects: From design to deployment. The section links point to the parts of the book it draws on.