Icon

News relevance - Training

There has been no title set for this workflow's metadata.

News relevance - Training

This workflow preprocesses labeled news articles, searches keywords and computes embeddings using the nodes of the spaCy extension. Next, it trains and applies two machine learning algorithms to classify news relevance, evaluates the performance and stores the best model.

ML training and testing with spaCy embeddings
Visualization
Model deployment
Text exploration, preprocessing and embeddings creation

News relevance - Training


This workflow preprocesses labeled news articles, searches keywords and computes embeddings using the nodes of the spaCy extension. Next, it trains and applies two machine learning algorithms to classify news relevance, evaluates the performance and stores the best model.

Data import
Label test records
XGBoost Predictor
en_core_web_md
Spacy Model Selector
Label test records
Decision Tree Predictor
Decision Tree Learner
Model evaluation
XGBoost Tree Ensemble Learner
Labellednewspaper articles
CSV Reader
Unwind embeddingsas columns
Split Collection Column
Color by relevance class
Color Manager
Co-occurenceTag cloud
Text analysis dashboard
Double-click toconfigure
Keyword Search
Label
Number To String
80% training 20% test
Partitioning
Save best model
Model Writer

Nodes

Extensions

Links