Searcharxiv⌕ Search

arXiv subjects

Selina Kurer

Publications and source records attributed to Selina Kurer.

2 recordsLinked to original sources
arXiv2024

Optimal multi-action treatment allocation: A two-phase field experiment to boost immigrant naturalization

Research underscores the role of naturalization in enhancing immigrants' socio-economic integration, yet application rates remain low. We estimate a policy rule for a letter-based information campaign encouraging newly eligible immigrants in Zurich, Switzerland, to naturalize. The policy rule assigns one out of three treatment letters to each individual, based on their observed characteristics. We field the policy rule to one-half of 1,717 immigrants, while sending random treatment letters to the other half. Despite only moderate treatment effect heterogeneity, the policy tree yields a larger, albeit insignificant, increase in application rates compared to assigning the same letter to everyone.

econ.GN↗
arXiv2023

Human-in-the-Loop Hate Speech Classification in a Multilingual Context

The shift of public debate to the digital sphere has been accompanied by a rise in online hate speech. While many promising approaches for hate speech classification have been proposed, studies often focus only on a single language, usually English, and do not address three key concerns: post-deployment performance, classifier maintenance and infrastructural limitations. In this paper, we introduce a new human-in-the-loop BERT-based hate speech classification pipeline and trace its development from initial data collection and annotation all the way to post-deployment. Our classifier, trained using data from our original corpus of over 422k examples, is specifically developed for the inherently multilingual setting of Switzerland and outperforms with its F1 score of 80.5 the currently best-performing BERT-based multilingual classifier by 5.8 F1 points in German and 3.6 F1 points in French. Our systematic evaluations over a 12-month period further highlight the vital importance of continuous, human-in-the-loop classifier maintenance to ensure robust hate speech classification post-deployment.

cs.CL↗