yumath / Bertner
ChineseNER based on BERT, with BiLSTM+CRF layer
Stars: ✭ 195
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Bertner
Turkish Bert Nlp Pipeline
Bert-base NLP pipeline for Turkish, Ner, Sentiment Analysis, Question Answering etc.
Stars: ✭ 85 (-56.41%)
Mutual labels: named-entity-recognition, ner
Persian Ner
پیکره بزرگ شناسایی موجودیتهای نامدار فارسی برچسب خورده
Stars: ✭ 183 (-6.15%)
Mutual labels: named-entity-recognition, ner
Bi Lstm Crf Ner Tf2.0
Named Entity Recognition (NER) task using Bi-LSTM-CRF model implemented in Tensorflow 2.0(tensorflow2.0 +)
Stars: ✭ 93 (-52.31%)
Mutual labels: named-entity-recognition, ner
Kashgari
Kashgari is a production-level NLP Transfer learning framework built on top of tf.keras for text-labeling and text-classification, includes Word2Vec, BERT, and GPT2 Language Embedding.
Stars: ✭ 2,235 (+1046.15%)
Mutual labels: named-entity-recognition, ner
Bert Sklearn
a sklearn wrapper for Google's BERT model
Stars: ✭ 182 (-6.67%)
Mutual labels: named-entity-recognition, ner
Torchcrf
An Inplementation of CRF (Conditional Random Fields) in PyTorch 1.0
Stars: ✭ 58 (-70.26%)
Mutual labels: named-entity-recognition, ner
Ner
命名体识别(NER)综述-论文-模型-代码(BiLSTM-CRF/BERT-CRF)-竞赛资源总结-随时更新
Stars: ✭ 118 (-39.49%)
Mutual labels: named-entity-recognition, ner
Named entity recognition
中文命名实体识别（包括多种模型：HMM，CRF，BiLSTM，BiLSTM+CRF的具体实现）
Stars: ✭ 995 (+410.26%)
Mutual labels: named-entity-recognition, ner
Ner Evaluation
An implementation of a full named-entity evaluation metrics based on SemEval'13 Task 9 - not at tag/token level but considering all the tokens that are part of the named-entity
Stars: ✭ 126 (-35.38%)
Mutual labels: named-entity-recognition, ner
Dan Jurafsky Chris Manning Nlp
My solution to the Natural Language Processing course made by Dan Jurafsky, Chris Manning in Winter 2012.
Stars: ✭ 124 (-36.41%)
Mutual labels: named-entity-recognition, ner
Phonlp
PhoNLP: A BERT-based multi-task learning toolkit for part-of-speech tagging, named entity recognition and dependency parsing (NAACL 2021)
Stars: ✭ 56 (-71.28%)
Mutual labels: named-entity-recognition, ner
Ncrfpp
NCRF++, a Neural Sequence Labeling Toolkit. Easy use to any sequence labeling tasks (e.g. NER, POS, Segmentation). It includes character LSTM/CNN, word LSTM/CNN and softmax/CRF components.
Stars: ✭ 1,767 (+806.15%)
Mutual labels: named-entity-recognition, ner
Ner blstm Crf
LSTM-CRF for NER with ConLL-2002 dataset
Stars: ✭ 51 (-73.85%)
Mutual labels: named-entity-recognition, ner
Sequence tagging
Named Entity Recognition (LSTM + CRF) - Tensorflow
Stars: ✭ 1,889 (+868.72%)
Mutual labels: named-entity-recognition, ner
Jointre
End-to-end neural relation extraction using deep biaffine attention (ECIR 2019)
Stars: ✭ 41 (-78.97%)
Mutual labels: named-entity-recognition, ner
Bond
BOND: BERT-Assisted Open-Domain Name Entity Recognition with Distant Supervision
Stars: ✭ 96 (-50.77%)
Mutual labels: named-entity-recognition, ner
Tf ner
Simple and Efficient Tensorflow implementations of NER models with tf.estimator and tf.data
Stars: ✭ 876 (+349.23%)
Mutual labels: named-entity-recognition, ner
Nlp Experiments In Pytorch
PyTorch repository for text categorization and NER experiments in Turkish and English.
Stars: ✭ 35 (-82.05%)
Mutual labels: named-entity-recognition, ner
Multilstm
keras attentional bi-LSTM-CRF for Joint NLU (slot-filling and intent detection) with ATIS
Stars: ✭ 122 (-37.44%)
Mutual labels: named-entity-recognition, ner
Bnlp
BNLP is a natural language processing toolkit for Bengali Language.
Stars: ✭ 127 (-34.87%)
Mutual labels: named-entity-recognition, ner
Bert-ChineseNER
Introduction
该项目是基于谷歌开源的BERT预训练模型，在中文NER任务上进行fine-tune。
Datasets & Model
训练本模型的主要标记数据，来自于zjy-usas的ChineseNER项目。本项目在原本的BiLSTM+CRF的框架前，添加了BERT模型作为embedding的特征获取层，预训练的中文BERT模型及代码来自于Google Research的bert。
Results
引入bert之后，可以看到在验证集上的F-1值在训练了16个epoch时就已经达到了94.87，并在测试集上达到了93.68，在这个数据集上的F-1值提升了两个多百分点。
Train
- 下载bert模型代码，放入本项目根目录
- 下载bert的中文预训练模型，解压放入本项目根目录
- 搭建依赖环境python3+tensorflow1.12
- 执行
python3 train.py即可训练模型 - 执行
python3 predict.py可以对单句进行测试
整理后的项目目录，应如图所示。
Conclusion
可以看到，使用bert以后，模型的精度提升了两个多百分点。并且，在后续测试过程中发现，使用bert训练的NER模型拥有更强的泛化性能，比如训练集中未见过的公司名称等，都可以很好的识别。而仅仅使用ChineseNER中提供的训练集，基于BiLSTM+CRF的框架训练得到的模型，基本上无法解决OOV问题。
Fine-tune
目前的代码是Feature Based的迁移，可以改为Fine-tune的迁移，效果还能再提升1个点左右。fine-tune可以自行修改代码，将model中的bert参数加入一起训练，并将lr修改到1e-5的量级。 并且，是否添加BiLSTM都对结果影响不大，可以直接使用BERT输出的结果进行解码，建议还是加一层CRF，强化标记间的转移规则。
Reference
(1) https://github.com/zjy-ucas/ChineseNER
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].