Developer friendly Natural Language Processing ✨

基于 TensorFlow & PaddlePaddle 的通用序列标注算法库（目前包含 BiLSTM+CRF, Stacked-BiLSTM+CRF 和 IDCNN+CRF，更多算法正在持续添加中）实现中文分词（Tokenizer / segmentation）、词性标注（Part Of Speech, POS）和命名实体识别（Named Entity Recognition, NER）等序列标注任务。

Extract city and country mentions from Text like GeoText without regex, but FlashText, a Aho-Corasick implementation.

Data for the HIPE 2022 shared task.

Tool for slot extraction from text

Contains a Keras Bi-LSTM for Named Entity Recognition (This example demonstrates how you can use Kubeflow to train and deploy a Keras model with a custom prediction routine).

Contains jupyter notebooks, presentations and examples for Keras, Google AI Platform and Kubeflow.

Project to extract entities from Job Description Articles.

Sample Recruitment Bot for Oracle Digital Assistant

Making a custom entity extraction model using spacy 3.5 using both conventional and transformer background. I will also try the spancat pipepline along with the ner.

A node binding for the MIT Information Extraction library.

NLP classification models built using intimate partner violence news articles

Successfully developed a Named Entity Recognition (NER) model on the BC5CDR dataset using Stacked Bidirectional GRUs with Attention mechanism, designed to accurately identify chemical and disease entities from biomedical texts.

Successfully developed a Named Entity Recognition (NER) model using a Bidirectional GRU with Attention on the MIT Movies dataset to identify and classify movie-related entities like titles, actors, and genres.

The tool for converting Edo era's Japanese business records (titles) into named entities. For this purpose, you need to make a user dictionary of Mecab.

Master Thesis Code

Documentation for Tisane

Successfully developed a Named Entity Recognition (NER) model for German text using a Bidirectional LSTM with Attention on the Multilingual NER dataset, effectively identifying entities across multilingual corpora with contextual understanding.