Fig. 3: Example results of applying a machine-assisted quantitizing approach to real-world data.

A Zero-shot prediction of predefined topics in the corpus of Soviet newsreel synopses. The vertical axis shows yearly aggregate percentages. Bootstrapped confidence intervals are added to the trend of the Social topic. There is less data in the latter years, reflected in the wider intervals. B Wrecking causes of ships found in the Baltic Sea, mostly in Estonian waters, as annotated by experts based on field notes and historical documents (left), compared to zero-shot prediction of said categories based on the same data, with bootstrapped confidence intervals on the counts. Due to fairly good classification accuracy, the counts end up roughly similar.