part_of_speech_tagging.md

Part of Speech Tagging

VLSP 2013

27,870 sentences for training and development from the VLSP 2013 POS tagging shared task:

• 27k sentences are used for training.
• 870 sentences are used for development.

Test data: 2120 test sentences from the VLSP 2013 POS tagging shared task.

Model	Accuracy	Method	Reference	Code
VnMarMoT	95.88	Nguyen et al. NAACL'18		Official
BiLSTM-CRFs + CNN-char	95.40	Ma et al. ACL'16	Nguyen et al. NAACL'18	Link
BiLSTM-CRF + LSTM-char	95.31	Lample et al. NAACL'16	Nguyen et al. NAACL'18	Link
BiLSTM-CRF	95.31	Huang et al. ArXiv'15	Nguyen et al. NAACL'18	Link
RDRPOSTagger	95.11	Nguyen et al. EACL'14		Official
JointWPD	94.03	Nguyen et al. '18

VietTreeBank

• 📜 VietTreeBank Paper

Dataset

• train: 7268 sentences, dev: 1038 sentences, test: 2077 sentences
• labels: N, V, CH, R, E, A, P, Np, M, N, Nc, L, T, Ny, Nu, X, B, S, I, Y, Vy

Model	Accuracy	Method	Reference	Code	Note
BiLSTM-CRFs	93.52	Nguyen et al. '18		Official	10-fold CV
VNTagger	93.40	Le et al. TALN'10		Official	10-fold CV
RDRPOSTagger	91.96		Pham et al. IJCNLP'17	Official	5-fold CV
NNVLP	91.92	Pham et al. IJCNLP'17		Official	5-fold CV
vTools	90.73	Tran et al. VLSP'13	Pham et al. IJCNLP'17	Official
Vitk	88.41		Pham et al. IJCNLP'17	Official

Miscellaneous

📜 Papers

• Nguyen et al. NICS'18. Building Vietnamese Linguistic Resources for Social Network Text Analysis
• Nguyen et al. ALTA'17, Nguyen et al. 2015
• Nguyen et al. 2014, Nguyen et al. 2011, Nguyen et al. 2011, Nguyen et al. 2010
• Tran et al. 2009, Phan et al. 2008, Nguyen et al. 2006
• Nguyen et al. 2003

💫 Services

• OpenFPT: Vietnamese Accentizer (2017)

📁 Open sources

• vncorenlp/VnCoreNLP (2018) java
• pth1993/NNVLP (2017) python,bash
• pyvi (2016) python
• Vitk (2016) java
• kanjirz50/viet-morphological-analysis-crf (2016), demonstration python
• lupanh/vTools (2015) python
• truongdo/vita (2015) c++
• RDRPOSTagger (2013-2017) python
• vnTagger (2010) java