Fig. 6
From: On the Readiness of Scientific Data Papers for a Fair and Transparent Use in Machine Learning
Evolution of the least informed dimensions from 2015 to 2023 (June). Figure (b) is evaluated only on datasets gathered from or describing people (16,5% of the sample). Figure (e) and Figure (f) are evaluated only on datasets collected from or representing people. Figure (f) is evaluated only on natural language datasets. Finally, Figures (g), (h), and (i) are evaluated only on datasets created trough an annotation process.
