A Unified MRC Framework for Named Entity Recognition

The repository contains the code of the recent research advances in Shannon.AI.

A Unified MRC Framework for Named Entity Recognition
Xiaoya Li, Jingrong Feng, Yuxian Meng, Qinghong Han, Fei Wu and Jiwei Li
In ACL 2020. paper
If you find this repo helpful, please cite the following:

@article{li2019unified,
  title={A Unified MRC Framework for Named Entity Recognition},
  author={Li, Xiaoya and Feng, Jingrong and Meng, Yuxian and Han, Qinghong and Wu, Fei and Li, Jiwei},
  journal={arXiv preprint arXiv:1910.11476},
  year={2019}
}

For any question, please feel free to contact xiaoya_li@shannonai.com or post Github issues.

Contents

1. Overview
2. Experimental Results on Flat/Nested NER Datasets
3. Dependencies
4. Data Preprocess
5. Training BERT MRC-NER Model
6. Evaluating the Trained Model
7. Descriptions of Directories
8. Contact

Overview

The task of NER is normally divided into nested NER and flat NER depending on whether named entities are nested or not. Instead of treating the task of NER as a sequence labeling problem, we propose to formulate it as a SQuAD-style machine reading comprehension (MRC) task.

For example, the task of assigning the [PER] label to "[Washington] was born into slavery on the farm of James Burroughs" is formalized as answering the question "Which person is mentioned in the text?".

By unifying flat and nested NER under an MRC framework, we're able to gain a huge improvement on both flat and nested NER datasets, which achives SOTA results.

We use MRC-NER to denote the proposed framework.
Here are some of the highlights:

1. MRC-NER works better than BERT-Tagger with less training data.
2. MRC-NER is capable of handling both flat and nested NER tasks under a unified framework.
3. MRC-NER has a better zero-shot learning ability which can predicts labels unseen from the training set.
4. The query encodes prior information about the entity category to extract and has the potential to disambiguate similar classes.

Experimental Results on Flat/Nested NER Datasets

Experiments are conducted on both Flat and Nested NER datasets. The proposed method achieves vast amount of performance boost over current SOTA models.
We only list comparisons between our proposed method and previous SOTA in terms of span-level micro-averaged F1-score here. For more comparisons and span-level micro Precision/Recall scores, please check out our paper.

Flat NER Datasets

Evaluations are conducted on the widely-used bechmarks: CoNLL2003, OntoNotes 5.0 for English; MSRA, OntoNotes 4.0 for Chinese. We achieve new SOTA results on OntoNotes 5.0, MSRA and OntoNotes 4.0, and comparable results on CoNLL2003.

Dataset	Eng-OntoNotes5.0	Zh-MSRA	Zh-OntoNotes4.0
Previous SOTA	89.16	95.54	80.62
Our method	91.11	95.75	82.11
	(+1.95)	(+0.21)	(+1.49)

Nested NER Datasets

Evaluations are conducted on the widely-used ACE 2004, ACE 2005, GENIA, KBP-2017 English datasets.

Dataset	ACE 2004	ACE 2005	GENIA	KBP-2017
Previous SOTA	84.7	84.33	78.31	74.60
Our method	85.98	86.88	83.75	80.97
	(+1.28)	(+2.55)	(+5.44)	(+6.37)

Previous SOTA:

• DYGIE for ACE 2004.
• Seq2Seq-BERT for ACE 2005 and GENIA.
• ARN for KBP2017.

Dependencies

• Experiments are conducted on a Ubuntu GPU server with Python 3.6.
  Run pip3 install -r requirements.txtto install packages dependencies.

• Download and unzip BERT-Base, Uncased English and BERT-Base, Chinese pretrained checkpoints. Then follow the guideline from huggingface to convert TF checkpoints to PyTorch.

Data Preprocess

Firstly you should transform tagging-style annoations to a set of MRC-style (Query, Context, Answer) triples. Here we provide an example to show how these two steps work. We have given the queries in python3 ./data_preprocess/dump_query2file.py for you. Feel free to write down your own's queries. MRC-Style datasets could be found here.

Step 1: Query Generation

Write down queries for entity labels in ./data_preprocess/dump_query2file.py and run python3 ./data_preprocess/dump_query2file.py to dump queries to the folder ./data_preprocess/queries.

Step 2: Transform tagger-style annotations to MRC-style triples

Run ./data_preprocess/example/generate_data.py to generate MRC-style data data_preprocess/example/mrc-dev_ace05.json and data_preprocess/example/mrc-dev_msra.json for ACE 2005(nested) and Chinese MSRA(flat), respectively.

Nested NER

We take ACE2005 as an example for NESTED NER to illustrate the process of data prepration.

Source files for ACE2005 contains a list of json in the format :

{
"context": "begala dr . palmisano , again , thanks for staying with us through the break .",
"label": {
    "PER": [
        "1;2",
        "1;4",
        "11;12"],
    "ORG": [
        "1,2"]
}
}

It assumes queries for ACE2005 should be found in ../data_preprocess/queries/en_ace05.json. The path for the queries should be registered in dictionary queries_for_dataset of ../data_preprocess/query_map.py.

Run the following commands to get MRC-style data files.

$ python3 
> from data_preprocess.generate_mrc_dataset import generate_query_ner_dataset
> source_file_path = "$PATH-TO-TAGGER-ACE05$/dev_ace05.json"
> target_file_path = "$PATH-TO-MRC-ACE05$/mrc-dev_ace05.json"
> entity_sign = "nested" #"nested" for nested-NER; "flat" for flat-NER.
> dataset_name = "en_ace2005" 
> query_sign = "default"
> generate_query_ner_dataset(source_file_path, target_file_path, entity_sign=entity_sign, dataset_name=dataset_name, query_sign=query_sign)

After that, $PATH-TO-MRC-ACE05$/mrc-dev_ace05.json contains a list of jsons:

{
"context": "begala dr . palmisano , again , thanks for staying with us through the break .",
"end_position": [
    2,
    4,
    12
    ],
"entity_label": "PER",
"impossible": false,
"qas_id": "4.3",
"query": "3",
"span_position": [
    "1;2",
    "1;4",
    "11;12"],
"start_position": [
    1,
    1,
    11]
}

Flat NER

Take Chinese MSRA as an example to illuatrate the process for FLAT NER.

Source files are in CoNLL format and entities are annotated with BMES scheme :

begala B-PER
dr M-PER
palmisano E-PER
, O
again O 
, O
thanks O
for O
staying O
with O
us O
through O
the O
break O
. O

Queries for Chinese MSRA should be found in ./data_preprocess/queries/zh_msra.json. The path for the queries should be registered in dictionary queries_for_dataset of ./data_preprocess/query_map.py.

Run the following commands to get MRC-style datasets:

$ python3 
> from data_preprocess.generate_mrc_dataset import generate_query_ner_dataset
> source_file_path = "$PATH-TO-TAGGER-ZhMSRA$/dev_msra.bmes"
> target_file_path = "$PATH-TO-MRC-ZhMSRA$/mrc-dev_msra.json"
> entity_sign = "flat" #"nested" for nested-NER; "flat" for flat-NER.
> dataset_name = "zh_msra" 
> query_sign = "default"
> generate_query_ner_dataset(source_file_path, target_file_path, entity_sign=entity_sign, dataset_name=dataset_name, query_sign=query_sign)

After that, $PATH-TO-MRC-ZhMSRA$/mrc-dev_msra.json contains a list of jsons:

{
"context": "begala dr . palmisano , again , thanks for staying with us through the break .",
"end_position": [2],
"entity_label": "PER",
"impossible": false,
"qas_id": "4.3",
"query": "3",
"span_position": [
    "1;2"],
"start_position": [1]
}

Training BERT MRC-NER Model

You can directly use the following commands to train the MRC-NER model with some minor changes.
data_sign should take the value of [conll03, zh_msra, zh_onto, en_onto, genia, ace2004, ace2005, kbp17, resume].
entity_sign should take the value of [flat, nested].

#!/usr/bin/env bash 
# -*- coding: utf-8 -*- 

FOLDER_PATH=/PATH-TO-REPO/mrc-for-flat-nested-ner
CONFIG_PATH=${FOLDER_PATH}/config/zh_bert.json
DATA_PATH=/PATH-TO-BERT_MRC-DATA/zh_ontonotes4
BERT_PATH=/PATH-TO-BERT-CHECKPOINTS/chinese_L-12_H-768_A-12
EXPORT_DIR=/PATH-TO-SAVE-MODEL-CKPT
data_sign=zh_onto 
entity_sign=flat

export PYTHONPATH=${FOLDER_PATH}
CUDA_VISIBLE_DEVICES=0 python3 ${FOLDER_PATH}/run/train_bert_mrc.py \
--config_path ${CONFIG_PATH} \
--data_dir ${DATA_PATH} \
--bert_model ${BERT_PATH} \
--output_dir ${EXPORT_DIR} \
--entity_sign ${entity_sign} \
--data_sign ${data_sign} \
--n_gpu 1 \
--export_model True \
--dropout 0.3 \
--checkpoint 600 \
--max_seq_length 100 \
--train_batch_size 16 \
--dev_batch_size 16 \
--test_batch_size 16 \
--learning_rate 8e-6 \
--weight_start 1.0 \
--weight_end 1.0 \
--weight_span 1.0 \
--num_train_epochs 10 \
--seed 2333 \
--warmup_proportion -1 \
--gradient_accumulation_steps 1

Evaluating the Trained Model

You can directly use the following commands to evaluate the MRC-NER model after training.
data_signshould take the value of [conll03, zh_msra, zh_onto, en_onto, genia, ace2004, ace2005, kbp17, resume].
entity_sign should take the value of [flat, nested].

#!/usr/bin/env bash 
# -*- coding: utf-8 -*- 

REPO_PATH=/PATH-TO-REPO/mrc-for-flat-nested-ner 
CONFIG_PATH=${FOLDER_PATH}/config/zh_bert.json
DATA_PATH=/PATH-TO-BERT_MRC-DATA/zh_ontonotes4
BERT_PATH=/PATH-TO-BERT-CHECKPOINTS/chinese_L-12_H-768_A-12
SAVED_MODEL_PATH=/PATH-TO-SAVED-MODEL-CKPT/bert_finetune_model.bin
data_sign=zh_onto 
entity_sign=flat

export PYTHONPATH=${REPO_PATH}
CUDA_VISIBLE_DEVICES=0 python3 ${REPO_PATH}/run/evaluate_mrc_ner.py \
--config_path ${CONFIG_PATH} \
--data_dir ${DATA_PATH} \
--bert_model ${BERT_PATH} \
--saved_model ${SAVED_MODEL_PATH} \
--max_seq_length 100 \
--test_batch_size 32 \
--data_sign ${data_sign} \
--entity_sign ${entity_sign} \
--n_gpu 1 \
--seed 2333

Descriptions of Directories

Name	Descriptions
log	A collection of training logs in experments.
script	Shell files help to reproduce our results.
data_preprocess	Files to generate MRC-NER train/dev/test datasets.
metric	Evaluation metrics for Flat/Nested NER.
model	An implementation of MRC-NER based on Pytorch.
layer	Components for building MRC-NER model.
data_loader	Funcs for loading MRC-style datasets.
run	Train / Evaluate MRC-NER models.
config	Config files for BERT models.

Contact

Feel free to discuss papers/code with us through issues/emails! xiaoya_li AT shannonai.com