Research on network security based on Prefix Span algorithm

In this research, the application of network traffic data for the improvement of network security using the Prefix Span algorithm is examined. Thus, the identified algorithm in this paper, called Prefix Span, which is used to mine frequent sequential patterns is used to find of the features that point to cyber threats among the network traffic. Using the CICIDS 2017 dataset, the effectiveness of the proposed algorithm is demonstrated on identifying several attacks on a system, including DDoS attack, SQL injection attempt, port scanning. Data preprocessing involves cleaning the data, feature selection, and sequence mining of feature data, followed by the use of algorithm from the SPMF library, and then performing analysis to gauge the algorithm’s accuracy in detecting the disease. Thus, evaluation values which include precision, recall, and F1-score indicate high efficiency of algorithm when it is used to distinguish between normal and malicious network activity. Overall, this research embellishes on features offered by data mining, namely the Prefix Span Algorithm, as a tool to strengthen network security and address current cyber threats. Possible future studies’ objectives include improvements in terms of scalability, parameter tuning, and taking into account the new types of threats in the sphere of computer cybersecurity to strengthen and develop effective network protection concepts.