Python Arabic Tweets sentiment classifier (85% score)

Among what the code include:

• (~ 1955 Arabic text Tweets)
  
  
• Tweet Pre-Processing function (Unicode blocks range method included).
  
  
• Naïve Bays Implementation ( including Log Prior, n_gram, undercount filter, scaler, polynomial features, and up to (9) classifiers for training comparison)
  
  
• Tensorflow approach (up to (11) activation functions and (7) optimizers for training comparison)
  
  
• Count Vectorizer approach (Also multiple classifiers for training review)
  
  
• Data pipeline Implementation
  
  
  

# code is ready on google Colab and all need files are linked to the cloud (Github)