From: Arabic text summarization using deep learning approach
References | Method | ROUGE F1 | Dataset |
---|---|---|---|
Rush et al. [9] | ABS | 30.88 | Gigaword |
ABS+ | 31.00 | ||
See et al. [11] | Pointer-generator | 36.44 | CNN/daily mail |
Pointer-generator + coverage model | 39.53 | ||
Gehrmann et al. [29] | Pointer-generator + coverage penalty | 39.12 | |
Bottom-up summarization | 41.22 | ||
(Ours) | Pointer-generator + length penalty | 44.23 | AHS |