Digital Evidence Security System Design Using Blockchain Technology

Abstract

Digital evidence plays an essential role in meeting the forensic need to uncover cybercrime and search for trace information of perpetrators. Digital evidence is vulnerable to system changes, human error, theft, deletion, and data manipulation, requiring security efforts to maintain authenticity. This study offers optimization of the chain of custody systems to maintain digital evidence integrity using authentication applications connected to the website server database. The design of the chain of custody system uses blockchain technology and K-means clustering algorithm. This research process consists of two stages. The first stage is the prototype of blockchain-based user access authentication applications. The second stage is the implementation of K-means clustering to determine the place of data storage according to its classification. The results of this study are the maximum security for blockchain-based chain of custody with the efficiency value of this application of 94.73% and the system load value of 0.223%. The total cost of deploying the application is 0.026702786 ETH. Based on this research can help to secure digital evidence information.

Analysis of Commercial Airplane Accidents Worldwide Using K-Means Clustering

“…K-means is an unsupervised learning clustering method used to analyze data by grouping data with similar characteristics into one cluster or group. Data with distinct characteristics are classified into different clusters or groups [10]. The K-means method requires the user to set three parameters: the number of K clusters to be used, cluster initialization, and system distance.…”

“…After testing the data for clustering, the Silhouette index and elbow methods can be used to determine how accurate they are. Sum Square Error (SSE) is a statistical method for calculating the total difference between the achieved and actual values [10].…”

“…Preparing raw data for use is an integral part of the preprocessing stage because the quality of the input data greatly influences the quality of the output data produced [10]. Several stages are conducted at this point, including data cleaning, integration, and reduction.…”

“…This is primarily due to the aim of this method, which is to uncover hidden insights and knowledge within the used data. In this study, K-means clustering is utilized to analyze data by grouping data with similar characteristics into one cluster or group, and data with distinct features into different clusters or groups [10]. The K-means method iteratively clusters the data until all data in a specific cluster does not migrate to another [11].…”