What is a problem definition?
- An agreement between the Data Scientists and the customer(s).
- It defines a set of goals that everyone agrees on before work begins.
- It ensures there is a common understanding.
- It should be stored within the SharePoint folder for reference.
- A problem definition template is available in the initiatives SharePoint folder.
As part of writing the problem definition, you will need to include:
- some background information about the area
- what it is that needs to be solved
- how you intend to solve the problem
- any limitations and caveats if they exist
- what tools and data you will use
- any information governance and ethical considerations
Tips for writing a problem definition
- For sensitive data containing personally identifiable information (PII), you will need to get approval from the Information Governance (IG) team before you start the work.
- Explain how you intend to use the data.
- Ensure the customer understands the aim by providing plenty of detail.
- Honestly assess time frames and relay these to the customer.
- Your critical friend and Data Science Lead can assist with this process.
For more information, please see the internal Data Science Wiki.
Throughout the process, your critical friend will review the problem definition to check:
- if you understood the customer’s request correctly
- if your plans to solve the problems are appropriate