Analyzing malware log files for internet access investigation using Hadoop

• Main Author: Mat Deli, Mohd. Sharudin
• Format: Thesis
• Language: English
• Published: 2017
• Subjects: QA76 Computer software
• Online Access: http://eprints.utm.my/id/eprint/78843/1/MohdSharudinMatMAIS2017.pdf; http://eprints.utm.my/id/eprint/78843/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:110301

On the Internet, malicious software (malware) is one of the most serious threats to system security. Major complex issues and problems on any software systems are frequently caused by malware. Malware can infect any computer software that has connection to Internet infrastructure. There are many types of malware and some of the popular malwares are botnet, trojans, viruses, spyware and adware. Internet users with lesser knowledge on the malware threats are susceptible to this issue. To protect and prevent the computer and internet users from exposing themselves towards malware attacks, identifying the attacks through investigating malware log file is an essential step to curb this threat. The log file exposes crucial information in identifying the malware, such as algorithm and functional characteristic, the network interaction between the source and the destination, and type of malware. By nature, the log file size is humongous and requires the investigation process to be executed on faster and stable platform such as big data environment. In this study, the authors had adopted Hadoop, an open source software framework to process and extract the information from the malware log files that obtains from university’s security equipment. The Python program was used for data transformation then analysis it in Hadoop simulation environment. The analysis includes assessing reduction of log files size, performance of execution time and data visualization using Microsoft Power BI (Business Intelligence). The results of log processing have reduced 50% of the original log file size, while the total execution time would not increase linearly with the size of the data. The information will be used for further prevention and protection from malware threats in university’s network.