What Makes An Expert? Reviewing How ML Researchers Define "Expert"

Mark Díaz,Angela DR Smith
2024-11-01
Abstract:Human experts are often engaged in the development of machine learning systems to collect and validate data, consult on algorithm development, and evaluate system performance. At the same time, who counts as an 'expert' and what constitutes 'expertise' is not always explicitly defined. In this work, we review 112 academic publications that explicitly reference 'expert' and 'expertise' and that describe the development of machine learning (ML) systems to survey how expertise is characterized and the role experts play. We find that expertise is often undefined and forms of knowledge outside of formal education and professional certification are rarely sought, which has implications for the kinds of knowledge that are recognized and legitimized in ML development. Moreover, we find that expert knowledge tends to be utilized in ways focused on mining textbook knowledge, such as through data annotation. We discuss the ways experts are engaged in ML development in relation to deskilling, the social construction of expertise, and implications for responsible AI development. We point to a need for reflection and specificity in justifications of domain expert engagement, both as a matter of documentation and reproducibility, as well as a matter of broadening the range of recognized expertise.
Machine Learning,Computers and Society
What problem does this paper attempt to address?
The problems that this paper attempts to solve are: in the development process of machine learning (ML) systems, the definitions of "expert" and "expertise" are not clear, and how to better utilize and recognize different forms of expert knowledge. Specifically, the authors explored the following issues by reviewing 112 academic publications: - **Definitions of experts and expertise**: In ML research, who is regarded as an "expert" and what constitutes "expertise" are not always clear or well - defined. Many studies fail to clearly define these concepts, which affects which knowledge is recognized and legitimized. - **Ways of using expert knowledge**: Expert knowledge is often used to mine textbook knowledge or discrete information that is easy to remember and reproduce (such as data labeling), while ignoring other forms of knowledge. - **Social construction and power relations**: The way experts participate in ML development involves social construction, deskilling, and power dynamics. The identity of an expert endows specific knowledge and experience with authority, but it may also bring about extractive and demeaning participation patterns. - **Responsible AI development**: In order to support responsible AI development, it is necessary to reflect on and clarify the reasons for the participation of domain experts, not only for documentation and reproducibility, but also for expanding the range of recognized expertise. ### Main findings 1. **Vague definitions**: Nearly half of the studies (51 papers) did not provide clear criteria to describe the experts or non - experts they contacted. 2. **Limited forms of knowledge**: Expert knowledge is usually concentrated on textbook knowledge or easily memorable information, while ignoring forms of knowledge outside of formal education and professional certification. 3. **Social and power relations in expert participation**: The way experts participate in ML development is affected by social and power relations, which may lead to deskilling and unfair power distribution. 4. **Impact on responsible AI development**: Clear and diverse definitions of experts are crucial for ensuring the fairness and responsibility of AI systems. ### Method The authors retrieved papers containing keywords such as "expert", "expertise", and "domain expert" from dblp.org through the methods of systematic literature review and thematic analysis, and carried out detailed coding and analysis on 112 related papers. The research focused on the roles and participation methods of experts and non - experts in different stages of ML development. ### Conclusion This paper emphasizes the importance of clearly and diversely defining experts in ML development to ensure that a wider range of knowledge forms are recognized and to promote more fair and responsible AI development practices.