Sociodemographic Bias in Language Models: A Survey and Forward Path
It addresses the problem of harmful bias in language models for researchers and practitioners, but is incremental as a survey.
This paper surveys research on sociodemographic bias in language models over the past decade, organizing it into a typology covering bias types, quantification, and debiasing techniques, and concludes with a checklist of open questions to guide future work.
Sociodemographic bias in language models (LMs) has the potential for harm when deployed in real-world settings. This paper presents a comprehensive survey of the past decade of research on sociodemographic bias in LMs, organized into a typology that facilitates examining the different aims: types of bias, quantifying bias, and debiasing techniques. We track the evolution of the latter two questions, then identify current trends and their limitations, as well as emerging techniques. To guide future research towards more effective and reliable solutions, and to help authors situate their work within this broad landscape, we conclude with a checklist of open questions.