EARTHQUAKE HAPPENS OR NOT USING DATASCIENCE TECHNIQUE

With the increasing popularity of online social networks, spammers find these medias easily accessible to trap users in malicious activities by posting spam messages. Here, Twitter platform is focused to analyze and perform the detection of spam tweets. To avoid the activities of spammers, Google Safe Browsing and Twitter’s Bot Maker tools detect and block spam tweets. Even though these tools can block malicious links they cannot protect the user from spammers’ activities in real-time as early as possible. Some of the approaches are only using user-based features while others are based on tweet based features only. There is no comprehensive solution which can enhance and consolidate tweet’s text information along with the user based features. For solving this issue, a framework is proposed which considers the user based and tweet based features for classifying the tweets. The advantage of using tweet text feature is that the spam tweets can be identified and detected even if the spammer creates and login with a new account, which was not possible only with the user and tweet based features. Estimation and evaluation of the entire process is performed using four different machine learning algorithms namely – SVM, KNN, ANN, RF and Naive Bayes. With the help of Neural Networks, an accuracy of 91.65% can be achieved which surpass the existing solution approximately by 18%.