Table 2 Datasets of each subtask.
From: A semantic union model for open domain Chinese knowledge base question answering
Task | Training set | Deving set | Testing set |
|---|---|---|---|
Entity Mention Recognition | 13,267 | 975 | 9870 |
Entity Disambiguation | 60,522 | 6724 | 36,219 |
Relation Matching | 132,388 | 14,709 | 102,589 |
Joint task of entity disambiguation relation matching | 337,065 | 37,481 | 320,579 |
