Figure - PMC

Skip to main content

View full-text article in PMC

. Author manuscript; available in PMC: 2022 Jan 1.

Published in final edited form as: J Biomed Inform. 2020 Dec 15;113:103665. doi: 10.1016/j.jbi.2020.103665

Figure 2. — Analogy completion task performance by various embeddings for (a) top 1 accuracy results and (b) top 3 accuracy results. Performance is separated by semantic categories. Significant differences in category performance according to McNemar’s test with continuity correction between Radiopaedia and WG embeddings of a given dimension. Significance is denoted * for BH adjusted-p <0.05, ** for BH adjusted-p< 0.01 and *** for BH adjusted-p < 0.001. No marking means no statistical significance. Abbreviations: WG = Wikipedia 2014 + Gigaword 5 embeddings, d = embedding dimensions, BH = Benjamini-Hochberg

Figure 2. — Analogy completion task performance by various embeddings for (a) top 1 accuracy results and (b) top 3 accuracy results. Performance is separated by semantic categories. Significant differences in category performance according to McNemar’s test with continuity correction between Radiopaedia and WG embeddings of a given dimension. Significance is denoted * for BH adjusted-p <0.05, ** for BH adjusted-p< 0.01 and *** for BH adjusted-p < 0.001. No marking means no statistical significance. Abbreviations: WG = Wikipedia 2014 + Gigaword 5 embeddings, d = embedding dimensions, BH = Benjamini-Hochberg