Skip to content
Sentiment analysis of tweets to U.S. airlines.
Branch: master
Clone or download
Will.Huguenin Will.Huguenin
Will.Huguenin and Will.Huguenin reorg and requirements file
Latest commit d17ca37 Jul 31, 2018

README.md

Metis Project 4: Rule-Base vs. ML-Trained Sentiment Analysis

How does the performance of various supervised ML and rule-based methods compare when tasked with identifying sentiment in Tweets to U.S. airlines?

Read my blog post for a full discussion, or check out the slides from my Metis presentation.

Program Flow

The table below provides high-level overviews of what each analysis script does. More information (including specific input/ouput data) can be found in each script's header.

Program Description
01-Apply-Sentiment.py Add rule-based TextBlob and VADER sentiment polarity scores to CrowdFlower data.
02-Text-Blob-Sentiment-Analysis.ipynb Sentiment analysis of 1,000 tweets to United Airlines using TextBlob.
03-Airline-Sentiment-Analysis.ipynb Compare performance of rule-based and supervised ML-based sentiment classifiers.

Code Dependencies

Before running any of these scripts, please make sure to have downloaded:

NumPy, Pandas, Sci-Kit Learn, Matplotlib, Seaborn, Pylab, re, NLTK (including their twitter samples corpus), TextBlob, vaderSentiment, Pickle, Cnfg, Os, TweePy (Twitter API Python Wrapper), MongoDB and PyMongo

You can’t perform that action at this time.