Han, QiweiSchmitz, Sophie Susanne2022-10-252022-10-252022-01-212021-12-16http://hdl.handle.net/10362/144990ESG controversies may have enormous consequences for an individual company, its customers, investors, and other stakeholders. The objective of this work is to identify companies at high risk of ESG controversy based on public ESG data. By using machine learning solutions, early indicators in ESG data can be identified that provide insight into how likely a company is to face an ESG controversy. By using Random Forest models, the proportion of companies with a controversy among the flagged companies can be increased by 93, 5.6and 4.3times for the Environmental, Social and Governance pillar, respectively.engSustainabilityMachine learningBusiness and data analyticsEsgRandom forestEsg controversymaster thesis203063929