Twitter emotion detection with spam recognition using machine learning algorithms
Pallavi Dhanve, Pooja Kshirsagar, Supriya Bhosle, Sidhdeswari Chaudhari, Laxman Devokate · Journal of Emerging Technologies and Innovative Research · 2020
The online social networks are a very large growth in the world today, but the attacks are more common, including one of the attacks is the attack of Twitter in this spammer spreading several malicious tweets that can take the form of links or hash tags in the website and online services, which are too harmful for real users. To prevent these attacks, training tweets are added and, moreover, these problems are solved by extracting 12 lightweight functions, like the age of the account, no. of followers, no. to follow, no. of tweets, no. of re-tweets, etc. For the transmission of spam detection from tweets, the discretization of a function is important for the performance of spam detection. There is a great truth in the system that includes a total of 600 public tweets based on the URL-based security tool. Spam detection primarily creates the classification model that includes binary classification and can also be solved using the automatic learning algorithm. Machine learning algorithms such as the Naive Bayesian classifier or the vector support machine classifier have informed the behavior of the models. The system reported the impact of data-related factors, such as the relationship between spam and non-spam, the size of training data and data sampling, and detection performance. The implemented system function is the detection of simple and variable tweets of spam over time. The system shows how spam detection is a major challenge and bridges the gap between performance appraisals and focuses primarily on data, features and patterns to identify the real user and inform the user of spam when providing the valuable response binary. The contribution work is to detect the tweets of emotions in real time, because the new tweets come in the form of sequences and use the updated training data set.