tailieunhanh - Multitask learning for biomedical named entity recognition with cross-sharing structure

Biomedical named entity recognition (BioNER) is a fundamental and essential task for biomedical literature mining, which affects the performance of downstream tasks. Most BioNER models rely on domain-specific features or hand-crafted rules, but extracting features from massive data requires much time and human efforts. To solve this, neural network models are used to automatically learn features. |

TÀI LIỆU LIÊN QUAN