Table 6 Comparative results of different methods for NLE-LP mining task
From: Using an analogical reasoning framework to infer language patterns for negative life events
Method | Within-category | Across-category | |||||
|---|---|---|---|---|---|---|---|
prec@5 | prec@10 | MRR | prec@5 | prec@10 | MRR | ||
Traditional induction | PPMI | – | – | – | 0.1296 | 0.1241 | 0.2589 |
HAL | 0.2034 | 0.1842 | 0.3921 | 0.2116 | 0.1872 | 0.3820 | |
Analogical Reasoning | CBOW + COSINE | 0.4904 | 0.4300 | 0.7287 | 0.4684 | 0.4129 | 0.7037 |
SG + COSINE | 0.4345 | 0.3824 | 0.6683 | 0.4215 | 0.3705 | 0.6683 | |
CBOW + COSMUL | 0.4272 | 0.3809 | 0.6654 | 0.4004 | 0.3589 | 0.6278 | |
SG + COSMUL | 0.3961 | 0.3516 | 0.6286 | 0.3751 | 0.3321 | 0.6197 | |