Skip to content

vmarar/PredictingParties

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 

Repository files navigation

PredictingParties

This project aims to be able to identify what political ideology an account on social media is representing or is aligned with based on their language used.

To approach this problem tweets will be analyzed using a series of data visualization techniques, and then will be used to create machine learning models to predict political affiliation. The language used by Democrats or Republicans are specific to their ideologies, therefore analyzing tweets written by political representatives will be a great asset to our machine learning models. Political representatives use very politically correct and specific jargon that is unique to their political parties lexicon.

Using methods such as Count Vectorizer and TF-IDF, a series of models were created to predict what party a tweet is leaning towards or identifies as. The models will then be compared based on accuracy and their confusion matrix to determine which one is most suitable for this task.

APPROACH :

  1. Sentiment Analysis
  2. StopWords, Stemming and Lemmatization
  3. TF-IDF
  4. CountVectorizer

Models Used:

  1. Naive Bayes Multinomial Classifier
  2. SVM Linear Kernel
  3. Decision Tree Regressor

Thank you!

About

Predicting Political Parties Using Twitter Data

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published