English NER in Flair (Ontonotes默认模型)

这是内置在Flair中的用于英文的18类NER模型(默认为Ontonotes模型)。

F1评分: 89.27 (Ontonotes)

预测18个标签:

基于 Flair embeddings 和LSTM-CRF模型。

演示: 如何在Flair中使用

需要: Flair (pip install flair)

from flair.data import Sentence
from flair.models import SequenceTagger

# load tagger
tagger = SequenceTagger.load("flair/ner-english-ontonotes")

# make example sentence
sentence = Sentence("On September 1st George Washington won 1 dollar.")

# predict NER tags
tagger.predict(sentence)

# print sentence
print(sentence)

# print predicted NER spans
print('The following NER tags are found:')
# iterate over entities and print
for entity in sentence.get_spans('ner'):
    print(entity)

运行以下脚本将得到如下输出:

Span [2,3]: "September 1st"   [− Labels: DATE (0.8824)]
Span [4,5]: "George Washington"   [− Labels: PERSON (0.9604)]
Span [7,8]: "1 dollar"   [− Labels: MONEY (0.9837)]

在句子 "On September 1st George Washington won 1 dollar" 中，出现了"September 1st"(被标记为日期)，"George Washington"(被标记为人物)和"1 dollar"(被标记为金钱)这三个实体。

训练: 用于训练该模型的脚本

以下Flair脚本用于训练该模型:

from flair.data import Corpus
from flair.datasets import ColumnCorpus
from flair.embeddings import WordEmbeddings, StackedEmbeddings, FlairEmbeddings

# 1. load the corpus (Ontonotes does not ship with Flair, you need to download and reformat into a column format yourself)
corpus: Corpus = ColumnCorpus(
                "resources/tasks/onto-ner",
                column_format={0: "text", 1: "pos", 2: "upos", 3: "ner"},
                tag_to_bioes="ner",
            )

# 2. what tag do we want to predict?
tag_type = 'ner'

# 3. make the tag dictionary from the corpus
tag_dictionary = corpus.make_tag_dictionary(tag_type=tag_type)

# 4. initialize each embedding we use
embedding_types = [

    # GloVe embeddings
    WordEmbeddings('en-crawl'),

    # contextual string embeddings, forward
    FlairEmbeddings('news-forward'),

    # contextual string embeddings, backward
    FlairEmbeddings('news-backward'),
]

# embedding stack consists of Flair and GloVe embeddings
embeddings = StackedEmbeddings(embeddings=embedding_types)

# 5. initialize sequence tagger
from flair.models import SequenceTagger

tagger = SequenceTagger(hidden_size=256,
                        embeddings=embeddings,
                        tag_dictionary=tag_dictionary,
                        tag_type=tag_type)

# 6. initialize trainer
from flair.trainers import ModelTrainer

trainer = ModelTrainer(tagger, corpus)

# 7. run training
trainer.train('resources/taggers/ner-english-ontonotes',
              train_with_dev=True,
              max_epochs=150)

引用

使用该模型时，请引用以下论文。

@inproceedings{akbik2018coling,
  title={Contextual String Embeddings for Sequence Labeling},
  author={Akbik, Alan and Blythe, Duncan and Vollgraf, Roland},
  booktitle = {{COLING} 2018, 27th International Conference on Computational Linguistics},
  pages     = {1638--1649},
  year      = {2018}
}

问题?

Flair问题跟踪器可从 here 获取。