Abstract: Today, social media platforms are being utilized by a gazillion of people which covers a vast variety of media. Among this, around 192 million active accounts are stated as Twitter users. This discovered an increasing number of bot accounts are problematic that spread misinformation and humor, and also promote unverified information which can adversely affect various issues. So in this paper, we will detect bots on Twitter using machine learning techniques. A web application where we can verify if the account is a bot or a genuine account. We analyze the dataset extracted from the Twitter API which consists of both human and bot accounts. We analyze the important features like tweets, likes, retweets, etc., which are required to provide us with good results. We use this data to train our model using machine learning methods Decision Trees and Random Forest. For linking our model with the web content, we used the flask server. Our result on our framework indicates that the user belongs to a human account or a bot with reasonable accuracy.
Keywords: Machine learning, Twitter, bot detection, Random forest.
|
DOI:
10.17148/IJARCCE.2021.10417
