kazemnejad / followers-comparison-in-twitter

Text classification of posts & contents created by followers of two celebrity in Twitter (code+data+slides)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Followers Comparison in Twitter data

Data collection & scraper scripts can be found on branch "master" Checkout results in report.pdf & report_slides.pdf

Single layer Convolution Neural network text classifier implementation in Tensorflow
*Check branch master for Naive Bayes model

Model Hyper Parameters:

  • Filter size: 2, 3, 4 windows
  • Number filters: 32
  • Word embedding size: 200
  • Maximum document length: 40
  • Static & dynamic word embedding is used
  • Static word vectors is trained using GloVe model on "52M+ token" corpus, check out results

About

Text classification of posts & contents created by followers of two celebrity in Twitter (code+data+slides)


Languages

Language:Python 100.0%