Abstract

Computer intrusion detection is essential for protecting information systems from malicious actions. With the exponential growth of data in modern computing environments, traditional ap- proaches to intrusion detection encounter significant difficulties in effectively managing and analyzing large-scale data. This paper investigates the use of big data technology to improve computer intrusion detection systems. The proposed approach seeks to improve the accuracy, scalability, and efficiency of intrusion detection by leveraging the capabilities of big data analytics, including distributed computing, parallel processing, and advanced machine learning al- gorithms. Using real-world datasets, a comprehensive evaluation is conducted to compare the performance of the optimized intrusion detection system with that of conventional methods. According to the findings of the experiment, the results of the classification algorithm are clearly superior to those of the clustering algorithm and the association rule. Even though there has been a rise in the number of atypical command percentages, the accuracy rate can still be kept at 92%. The association rule algorithm’s accuracy rate is basically maintained at above 81% as a balance between the other two algorithms. It has been demonstrated that the application of big data techniques to the problem of data or network intrusion detection can significantly enhance the ability to detect instances of computer network infiltration.