Skip to content

Demonstrates how careful selection can be more useful than including everything when making a punctuation feature.

Notifications You must be signed in to change notification settings

jcarterlab/NLP-punctuation-feature-engineering

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 

Repository files navigation

NLP-punctuation-feature-engineering

This notebook looks at how to optimize punctuation feature engineering for a Kaggle essay scoring competition. It turns out that using a count of all punctuation types can reduce effectiveness by adding unnecessary noise. In this case, using only the 3 types of punctuation with the strongest correlation to essays scores proved most effective.

The notebook can be found here.

About

Demonstrates how careful selection can be more useful than including everything when making a punctuation feature.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published