CAN-NER: Convolutional Attention Network for Chinese Named Entity Recognition

NAACL 2019  ยท  Yuying Zhu, Guoxin Wang, Bรถrje F. Karlsson ยท

Named entity recognition (NER) in Chinese is essential but difficult because of the lack of natural delimiters. Therefore, Chinese Word Segmentation (CWS) is usually considered as the first step for Chinese NER. However, models based on word-level embeddings and lexicon features often suffer from segmentation errors and out-of-vocabulary (OOV) words. In this paper, we investigate a Convolutional Attention Network called CAN for Chinese NER, which consists of a character-based convolutional neural network (CNN) with local-attention layer and a gated recurrent unit (GRU) with global self-attention layer to capture the information from adjacent characters and sentence contexts. Also, compared to other models, not depending on any external resources like lexicons and employing small size of char embeddings make our model more practical. Extensive experimental results show that our approach outperforms state-of-the-art methods without word embedding and external lexicon resources on different domain datasets including Weibo, MSRA and Chinese Resume NER dataset.

PDF Abstract NAACL 2019 PDF NAACL 2019 Abstract

Results from the Paper


Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Chinese Named Entity Recognition MSRA CAN-NER Model F1 92.97 # 20
Precision 93.53 # 3
Recall 92.42 # 3
Chinese Named Entity Recognition OntoNotes 4 CAN-NER Model F1 73.64 # 15
Precision 75.05 # 4
Recall 72.29 # 4
Chinese Named Entity Recognition Resume NER CAN-NER Model F1 94.94 # 12
Precision 95.05 # 3
Recall 94.82 # 3
Chinese Named Entity Recognition Weibo NER CAN-NER Model Accuracy-NE 55.38 # 1
Accuracy-NM 62.98 # 1
Overall 59.31 # 1

Methods


No methods listed for this paper. Add relevant methods here