International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 05 | May 2023
p-ISSN: 2395-0072
www.irjet.net
Detecting Malicious Bots in Social Media Accounts Using Machine Learning Technology. Megha Choudhar1, Pallavi Garje2, Deeksha Shinde3, Rutika Valanj4 1Mr. Dattatray Modani Professor, Dept. of Computer Engineering, P.E.S. Modern College of Engineering, Pune,
Maharashtra,India
1234UG Students, Dept. of Computer Engineering, P.E.S. Modern College of Engineering, Pune, Maharashtra, India
---------------------------------------------------------------------***--------------------------------------------------------------------an activity as benign or malicious. This approach offers several advantages over traditional methods, of artificial intelligence and machine learning in everyday life is increasing rapidly. Malicious bot activity is a including greater accuracy, efficiency, and scalability. significant threat to the security and integrity of online Furthermore, machine learning-based systems can services and systems. Machine learning technology provides adapt and learn over time, improving their a powerful tool for detecting and preventing such activity by performance and effectiveness. Despite the promise of identifying patterns and characteristics of bot behavior. This machine learning-based solutions, there are several approach involves collecting large data sets of both benign challenges and considerations that researchers and and malicious bot activities, extracting relevant features, practitioners must address. These include feature selecting appropriate classification algorithms, and selection, data preprocessing, algorithm selection, and continuously monitoring the model in real time. Unique real-time monitoring. Moreover, the ethical and legal concepts such as feature importance, correlation analysis, implications of deploying machine learning for bot domain knowledge, recursive feature elimination, wrapper methods, and principal component analysis can be used for detection must be carefully considered, including feature selection. The choice of feature selection and issues related to bias, privacy, and accountability. In classification algorithm will depend on the specific use case, this paper, we review the current state of the art in the characteristics of the data, and the desired accuracy and malicious bot detection using machine learning efficiency of the model. Using machine learning technology technology. We explore the various methods, to detect malicious bots provides a robust and effective algorithms, and techniques employed in this field, and solution to protect online services and systems against this discuss the benefits and limitations of each. We also growing threat. highlight some of the key challenges and considerations that must be addressed to develop Key Words: Social Media, Support Vector Machine (SVM), Pattern recognition, Anomaly Detection, effective and ethical machine learning-based solutions Decision Tree. for bot detection. Overall, we argue that machine learning has the potential to significantly improve the 1. INTRODUCTION security and integrity of online services and systems, and we call for continued research and innovation in TThe internet has transform the way we live, work, this important area. and communicate. However, with these benefits come several challenges, one of which is the growing 2. LITERATURE SURVEY threat of malicious bots. These automated software programs can infiltrate websites, social media 1. Detecting Malicious Twitter Bots Using platforms, and online services, causing damage to Behavioral Modeling and Machine Learning businesses and individuals alike. In response, Techniques by Liu et al. (2019) - This paper researchers and security professionals have proposes a method for detecting malicious developed a range of detection methods, including Twitter bots using a combination of behavioral traditional rule-based systems and more advanced modeling and ML techniques. The authors machine learning-based approaches. Machinecollect data on a set of Twitter accounts and learning technology has emerged as a powerful tool use features such as posting behavior, for detecting and preventing malicious bot activity. sentiment analysis, and network analysis to By leveraging algorithms and statistical models, train a set of classifiers, which are then used to machine learning can analyze large datasets of bot detect malicious bots with high accuracy. behavior, identify patterns and features, and classify
Abstract - In this age of technology boom, the application
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 173