Big data, algoritmer og behavioral public policy

This dissertation concerns the use of behavioral public policy, big data, and machine learning to address the problem of information overload in the labor market. Concretely, the dissertation develops and tests collaborative filtering and content-based recommender systems on the basis of government registry data for the purpose of providing personalized occupational recommendations that can assist in the job search process. The dissertation sheds light on the strengths and weaknesses of different recommender system approaches and data, with an emphasis on issues crucial for the public sector, such as algorithmic transparency and fairness.