Skip to content

An exercise in training and optimizing machine learning models to predict wether a product review (free text) will be helpful or unhelpful for other potential buyers.

Notifications You must be signed in to change notification settings

tfayemi/NLP_ReviewHelpfulnessClassifier

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 

Repository files navigation

NLP_ReviewHelpfulnessClassifier

Please veiw full report for complete breakdown of the models, how they work, and the intended results.

A collection of machine learning models trained on the Amazon's 2018 Review Data to predict wether a product review (free text) will be helpful or unhelpful for other potential buyers.

Pre-Step:

Download Stanford's GloVe 100d word embeddings (glove.6B.100d.txt) and store it in the same /helpful directory.

How To Use:

  1. Download the ‘helpful’ package and store it wherever you’d like.

  2. From your terminal or cmd line, run $ python3 helpful_api.py

  3. In another terminal/cmd line window, send you API requests in the following format (I didn’t have time to make a GUI unfortunately!)

$ curl -X POST -F "review=your review here" http://127.0.0.1:5000/predict

Replace “your review here” with your review!

And you should receive a JSON response to your terminal that looks something like this:

Example

HOPE IT WORKS! :)

About

An exercise in training and optimizing machine learning models to predict wether a product review (free text) will be helpful or unhelpful for other potential buyers.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages