Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.

Understanding User Comments via Sentiment Analysis

Includes analysis of a large corpus of positive and negative user comments, data cleaning, model selection, and deployment to a Flask REST API

Technologies: TensorFlow, Sklearn, NLTK, Pandas, Flask, Python


Install

packages:

  • pip3 install pickle-mixin
  • pip3 install tensorflow
  • pip3 install -U scikit-learn
  • pip3 install Flask
  • pip3 install nltk
  • pip3 install pandas

run:

  • train a model: (from the root folder) python comment_clf_model.py
  • run the server: (from the root folder) python comment_clf_app.py
  • go to the server home: http://127.0.0.1:5000/v1/api

Overview

This project represents a series of machine learning models used to identify attacks on users on Wikipedia using natural language processing. Using Scikit-learn and other packages, I built several classifiers that were able to predict whether a comment was an attack or not with a high rate of accuracy.


Notes

About

Understanding user comments via natural language processing with TensorFlow and Scikit-Learn

Resources

Releases

No releases published

Packages

No packages published