Table 3.
Effects of different prior knowledge on performance on the CDR dataset
| Method | Intra-sentence level | Inter-sentence level | Document level | ||||||
|---|---|---|---|---|---|---|---|---|---|
| P (%) | R (%) | F (%) | P (%) | R (%) | F (%) | P (%) | R (%) | F (%) | |
| KCN | 70.61 | 60.41 | 65.12 | 65.37 | 12.57 | 21.09 | 69.65 | 72.98 | 71.28 |
| AE | 71.44 | 57.04 | 63.43 | 57.71 | 10.88 | 18.31†† | 68.82 | 67.92 | 68.37† |
| SA | 60.99 | 53.38 | 56.93†† | 40.57 | 8.07 | 13.46†† | 57.21 | 61.44 | 59.25†† |
| AE-SA | 58.03 | 53.56 | 55.71†† | 47.69 | 5.82 | 10.37†† | 56.82 | 59.38 | 58.07†† |
The descriptions and analysis for Table 3 could be found in subsection “Effects of prior knowledge”. The marker † and †† represent P-value < 0.05 and P-value < 0.01, respectively, using pairwise t-test against KCN. The highest scores are highlighted in bold