Improve reliability and introduce model-based working methods in SSYK and SUN

Important results from the project

The project has mapped and evaluated the register production processes for the Swedish UREG and YREG with regard to classification by codes in the classifications SUN and SSYK. These activities aim to enable improvements in the processes together with an activity on modeling for continuous measurement of reliability in the coding of SUN and SSYK. The goal of the project is also for the work with UREG and YREG to become more automated, and to use opportunities with new technology such as AI/ML and to make greater use of already produced register data.

Expected long term effects

Expected results are mainly to simplify, improve and make visible the quality of the classifications SUN, SSYK in the population registers UREG and YREG in an automated way, as well as simplify and improve the work process with the production of UREG and YREG. The result of completed activities in this project should be able to have the following effects: - Potentials for improvement to increase quality in the long term - The improvement of the coding leads to safer decision-making bases - More automated process with new technology and new conditions means that the cost can be reduced

Approach and implementation

Involved partners in the project are the Employment Service, the Swedish School Agency and the Swedish University of Applied Sciences, as well as Statistics Sweden (project leading public authority). Identified competencies of the participants in the project are project management, methodologists, AI/ML competence, analyst and subject competence.The project was divided into four temporal phases 2023; Phase 1: Feb-Apr, Phase 2: May-Jun, Phase 3: Sep-Oct and Phase 4: Nov-Dec. - 20 project meetings - 8 main activities - 36 sub-activities