CYBER CRIME ESTIMATION AND IDENTIFICATION: A HYBRID KEY INDICATOR USING MACHINE LEARNING AND DATA ANALYTICS

CYBER CRIME ESTIMATION AND IDENTIFICATION: A HYBRID KEY INDICATOR USING MACHINE LEARNING AND DATA ANALYTICS

Mrs. Peddaboina Yamuna*1, Saideep Yamsani*2, Tejaswini Kachagandi*3,Vishal Chokkarapu.*4, Sagarika Posina*5

*1 Assistant Professor, Department Of CSE, ACE Engineering College, Hyderabad, Telangana, India.

*2,3,4,5Student, Department Of CSE, ACE Engineering College, Hyderabad, Telangana, India.

ABSTRACT
Social media text analytics involves extracting information from text sources, including social media, through text analysis. Social media plays a significant role in addressing the problem of crime, which is a major issue in society today. The occurrence of crime negatively impacts both quality of life and economic growth. By analyzing historical data, we can identify crime patterns and make predictions about future incidents. However, there are certain unreported and unsolved crimes due to insufficient evidence, making the task of locating criminals challenging. One way to address this is by monitoring social media for criminal activity, as users sometimes share information about their surroundings on these platforms. In this research paper, we propose a Machine Learning approach to detect crimes, analyze their types, and investigate/trace the evidence of the attacker. Initially, we retrieve text messages that contain predefined keywords related to crimes. Then, we pre-process the data and employ a Support Vector Machine-based filtering approach to eliminate noise. Random Forest is subsequently used for classification. Finally, in the last stage, we analyze and categorize the crime types. By using the categorized types, we identify the evidence associated with the attacker.

Keywords: Cyber Crime, Cyber Attacks, Security, SVM, Malware, Data Analytics, Machine Learning.