FLAT: Chinese NER Using Flat-Lattice Transformer

24 Apr 2020Xiaonan LiHang YanXipeng QiuXuanjing Huang

Recently, the character-word lattice structure has been proved to be effective for Chinese named entity recognition (NER) by incorporating the word information. However, since the lattice structure is complex and dynamic, most existing lattice-based models are hard to fully utilize the parallel computation of GPUs and usually have a low inference-speed... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Chinese Named Entity Recognition MSRA FLAT F1 94.12 # 7
Chinese Named Entity Recognition MSRA FLAT+BERT F1 96.09 # 2
Chinese Named Entity Recognition OntoNotes 4 FLAT F1 76.45 # 4
Chinese Named Entity Recognition OntoNotes 4 FLAT+BERT F1 81.82 # 3
Chinese Named Entity Recognition Resume NER FLAT F1 95.45 # 3
Chinese Named Entity Recognition Resume NER FLAT+BERT F1 95.86 # 2
Chinese Named Entity Recognition Weibo NER FLAT+BERT F1 68.55 # 1
Chinese Named Entity Recognition Weibo NER FLAT F1 60.32 # 3