#SASchat
A1: Data is not inherently biased. People are. Providing Data Governance oversight from a diverse group of decision-makers can help.
#saschat#trustworthyAI#analytics
A5: Data Governance is key to Data Management outcomes which in turn impacts model outcomes. Ensure there is a program objective regarding responsible innovation then actively monitor and measure (new responsibility for the Data Steward).
#saschat#trustworthyAI#analytics
A4: This is a great addition to a Data Steward’s list of responsibilities – someone should always be asking the question!
#saschat#trustworthyAI#analytics
A3: Set a good foundation for data management practices that includes defined common terms so AI doesn’t start with a model that is limited from the beginning. The more diverse the decision-makers, the better chance to reduce inequity.
#saschat#trustworthyAI#analytics
A2: Trustworthy means decision-makers trust the outputs. Solid Data Management methods and practices help with that. Data Governance can then provide the oversight and discipline to assure trusted outcomes.
#SASchat#trustworthyAI#analytics
A5: Develop data strategy with focus on Identify, Provision, Store, Integrate, Govern. The strategy should determine resource needs.
5 Components of a Data Strategy #saschat#government#analytics
A4: More integrated data solutions will provide better insights into programs - provide transparency and improve outcomes. #saschat#government#analytics
A3 - I see a lot of agencies with a lot of stand-alone applications which results in lots of one-off reporting applications - data integration is key! #saschat#government#analytics