AI to encourage data-driven decision making in public sector

Summary:

To promote data-driven decision-making in the public sector, the University of Latvia and Microsoft Innovation Center developed a Data Challenge platform. This initiative brings together teams of public sector officials and data analysts to create several AI solution prototypes within a few weeks, addressing specific public sector challenges. This platform enables organizations to explore and eventually implement innovative, data-driven solutions.

Client: 

The public sector organizations

Problem Statement: 

Data-driven problem solving and decision-making pose significant challenges in public administration. To wisely invest in new ideas, institutions need assurance that proposed solutions are effective, yet they often lack the opportunity to test these solutions beforehand.

Results: 

• This initiative allows public sector organizations to see and evaluate prototypes before committing to further development and investment.
• Refined operations and streamlined workflow with greater accuracy.
• Enhanced efficiency, accuracy, and cost-effectiveness by reducing human errors.

AI Solution Overview:

The University of Latvia and the Microsoft Innovation Center have developed the “Data Challenge” platform and methodology to help public sector organizations evaluate investment in innovation. This initiative organizes seminars for civil servants to tackle a specific, well-defined problem faced by public sector institutions. Data analytics teams then explore various data analysis techniques, from interactive data mining to machine learning and neural networks, to devise potential solutions.

Over a few weeks, teams of data analysts, technology experts, and enthusiasts analyze available datasets and other relevant information to prototype solutions. These teams present their ideas and working prototypes at a subsequent event, where data analysis experts assess the technical aspects, and organization representatives provide feedback on the practical applicability and usefulness of the solutions.

Various data analytics tools were employed to create an AI-based model for the judicial system in Latvia. This model could forecast and ascertain the probable duration of a trial. Utilizing the insights gained from this data-driven testing, they could refine their operations and streamline their workflow with greater accuracy.

The Rural Support Service in Latvia aimed to enhance the categorization process of images captured by local landowners and their staff to evaluate land conditions, specifically whether grass had been cut and removed. Teams participating in the “Data challenge” employed image recognition technology to address this task. Adopting such a solution promises to enhance efficiency, accuracy, and cost-effectiveness by reducing human errors. This will expedite the provision of government support to landowners, thus enhancing their cash flow.

References: 

1. “Data challenge” platform and methodology. https://oecd-opsi.org/innovations/data-challenge-platform-and-methodology/

Industry: Public Services

Vendors: University of Latvia and Microsoft Innovation Center 

Clients: Public sector organisations

Publication Date: 2017