news-inspector
News-inspector is a free software library for analyzing and acquiring information from news. It features various NLP and machine learning tasks that can be performed on text news articles:
- classification
- named entity recognition
- knowledge-based article search
All the methods can be easily configured and trained/retrained. Once trained, they can be loaded and used on new articles.
Requirements
- Python >=3.4
- scikit-learn>=0.17.1
- sklearn-crfsuite>=0.3
- nltk>=3.24
Installation
pip install news-inspector
Quick start
Train the model:
from news_inspector import GenericClassifier
train_model(GenericClassifier, "myconfig.xml", "myclassifier.model");
Load and use the model:
from news_inspector import load_model
model = load_model("myclassifier.model");
result = model.classify(text);
Documentation
Read the docs at https://news-inspector.readthedocs.io/en/latest/.