SPAM DETECTION IN TWITTER BY USING K-NEAREST NEIGHBOR (KNN)

Abstract

Since more than 300 million monthly users send 500 million tweets daily, Twitter is a popular social networking platform. This is the main reason why spammers use Twitter to spread malicious software that steals user personal information, tweets with faulty or fake URLs, assertively following or un-following users, trending fake tweets to attract users' attention, and spreading pornographic advertisements and among other reprehensible activities. The research clearly demonstrates that over 32 million people have engaged with the server for casual information on a daily basis. Twitter is said to have collected data on active users in previous years and studied their actions. Therefore, today's social media landscape, it is crucial to recognize and filter out the damaging or unwanted trends or malicious tweets. This technique suggests analyzing tweets and categorizing them as spam or ham based on the words they include. While there are several machine learning and deep learning techniques for categorizing and detecting spam tweets, this system will employ the clustering and binary detection model from KNN. This system is implemented using ASP. Net programming language with Microsoft SQL Server Database Engine.