Research Article
Multigranularity Pruning Model for Subject Recognition Task under Knowledge Base Question Answering When General Models Fail
Table 3
Experimental results of MGPM with different parameter values.
| | | | Accuracy (%) | Candidate |
| | 0.8 | 1 | 1 | 39.2 | <0.1k | | 0.7 | 1 | 1 | 46.6 | 0.5k | | 0.6 | 1 | 1 | 49.5 | 2.5k | | 0.5 | 1 | 1 | 49.5 | 8.6k | | 0.6 | 1 | 0.8 | 55.9 | 2.5k | | 0.6 | 1 | 0.5 | 56.8 | 2.5k | | 0.6 | 0.8 | 0.8 | 57.4 | 2.8k | | 0.6 | 0.8 | 0.5 | 58.3 | 2.9k | | 0.6 | 0.5 | 0.8 | 58.3 | 3.2k | | 0.6 | 0.5 | 0.5 | 59.4 | 3.4k | | 0.6 | 0.5 | 0 | 55.6 | 7.9k | | 0.6 | 0 | 0 | — | 109k | | 0 | 0 | 0 | — | 3,972k |
|
|
Bold values represent the best-performance.
|
