yumeng5/RoSTER
[EMNLP 2021] Distantly-Supervised Named Entity Recognition with Noise-Robust Learning and Language Model Augmented Self-Training
This project helps natural language processing researchers train models to identify specific entities like names or locations in text, even when the initial training data has errors or 'noise.' You provide raw text documents and a corresponding set of 'distant labels' (automatically generated or imperfect labels). The output is a highly accurate named entity recognition (NER) model, ready to be deployed or further refined.
No commits in the last 6 months.
Use this if you need to build a robust Named Entity Recognition model for your text data but only have access to distantly supervised or noisy training labels.
Not ideal if you have a perfectly clean, manually annotated dataset for your NER task, as its specialized noise-robust features would be overkill.
Stars
65
Forks
7
Language
Python
License
Apache-2.0
Category
Last pushed
Nov 12, 2021
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/yumeng5/RoSTER"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
hellohaptik/chatbot_ner
chatbot_ner: Named Entity Recognition for chatbots.
openeventdata/mordecai
Full text geoparsing as a Python library
Rostlab/nalaf
NLP framework in python for entity recognition and relationship extraction
mpuig/spacy-lookup
Named Entity Recognition based on dictionaries
NorskRegnesentral/skweak
skweak: A software toolkit for weak supervision applied to NLP tasks