Skip to content

Latest commit

 

History

History
41 lines (29 loc) · 858 Bytes

File metadata and controls

41 lines (29 loc) · 858 Bytes

Fake account detection in Twitter

Features

  • Uses machine learning algorithms to classify user accounts into real or fake.
  • Classifiers implemented are:
    • Naive Bayes
    • SVC
    • K-Nearest neighbour
  • Calcualtes accuracy and error rate of classifiers.
  • Displays comparison results of each classifier using bar graph.

A glimpse of running the model

Output Window

Images

Custom User Account

Images

Sample Classifier accuracy

Images

Classifier comparison

Images

How to run?

  • Install required libraries using
	pip install [library-name]
  • Execute the Python code using
	python model.py
  • Select the dataset "twitter_data.csv" or enter custom account details.
  • Choose the classifier and results will be displayed in new tab.