Predictive Analytics-Based Cybersecurity Framework for Cloud Infrastructure

Predictive Analytics-Based Cybersecurity Framework for Cloud Infrastructure

Akashdeep Bhardwaj, Keshav Kaushik
Copyright: © 2022 |Pages: 20
DOI: 10.4018/IJCAC.297106
(Individual Articles)
No Current Special Offers


The most valuable asset for any organization and individual is data and the information it holds. This is the main reason for Information Security to be the top concern in boardrooms and executive meetings. Security failures and data breaches now can impact an organization or a country's budget economy. To reduce Cybersecurity risks and improve data protection, there is an urgent need to implement a standard Framework for Cybersecurity. This framework utilizes AI and ML by including Policies, Guidelines, Standards and Practices, and data sources from Cloud Infrastructure systems like networks, servers, security systems, and end-user devices. Combining the data set gathered and risk governance information with Artificial Intelligence and Machine Learning. This research presents a framework that collects datasets, enriches and validates logs and datasets, then correlates them to analyze and predict the response to Cyber attack with high level of accuracy using ML model.
Article Preview

1. Introduction

Cybercriminals and Hackers target organizations and corporate businesses globally and business operations exposed to the highest levels of threats. Apart from discussing the business decisions and dynamic market demands, business owners are now discussing Cybersecurity Framework and the level at which the implementation has been successful for the business. Cybersecurity Frameworks help address cyber risks with two specific goals – the first is how to fight back cyberattacks, and the second, more importantly, how to mitigate the attacks altogether. New York Times recently reported an attack on US oil and gas companies where the hackers were rebuffed by the protection system set in place. They changed their approach to finding a more indirect fault and infected the website and menu of the local Chinese restaurant favored by the company's employees with malware. Their strategy paid off, and they were able to breach the company's network. This example, known as a Water Hole strategy, illustrates the level of complexity in addressing cybersecurity. Hackers are constantly changing their tactics and are always trying to stay ahead of the game. It is critical then to implement an efficient reporting system and keep Chief Information Security Officers (CISOs) updated on the latest types of attack to reduce reaction times to a minimum. On average it takes 46 days and $639,462 to resolve a cyberattack as per Predictive Analytics and the Future of Cybersecurity – Converge (2021). This is a statistic only for determined attacks, as this average doesn't include those that are dormant or undetected (Modern Day Attack).

A security breach can be detected using predictive analytics before it occurs. Like a radar that displays where and when an adversary is arriving, sophisticated analytics indicate when and where attacks may happen. This allows your business to sound the alert, raise the drawbridge, and get your soldiers ready. Predictive analytics will enable you to outsmart hackers and emerge triumphantly rather than finding a compromise after the fight has already been lost. The IT team is probably working full-time to develop creative, intelligent solutions as the number and sophistication of intrusions grow. The computer community has been looking for methods to keep sensitive information out of the hands of evil people. Emerging assaults, however, randomize their signatures, making them nearly challenging to identify and fight against. Predictive analytics has the potential to assist organizations in proactively identifying security problems before they cause harm. Companies can predict future occurrences and optimize prevention by concentrating on the “infection stage” of an assault rather than just the “infection stage.” Hacker bots utilize complicated analytics and extensive data to sniff out weaknesses before an exploit, similar to how IBM's mobile analyzer finds flaws in mobile apps.

Predictive analytics and hacker bots continuously monitor activities and provide essential information using self-learning analytics and detection approaches. It allows an organization to identify threats despite knowing the assault's precise signature, addressing a coverage gap that is currently ineffective against newer point-and-click vulnerabilities with unique attack signatures. Predictive analytics can spot anomalies in traffic flow and data right away, alerting you to a security concern before it happens. Predictive analytics is being used in a variety of sectors. As a result, predictive analytics' role in assisting organizations in identifying security vulnerabilities has risen to prominence. Quantzig (2021) reportedly conducted a poll to help evaluate cybersecurity challenges, and the results indicated that more than 70% of businesses are experiencing significant financial losses due to security breaches. In [3], all of the procedures involved in developing our metric security methodology are captured in the Cyber Security Analytics Framework. The framework is enhanced by the inclusion of temporal elements related to individual vulnerabilities. Researchers can forecast how the absolute security of the network evolves by utilizing Attack Graphs to capture their interdependence. The time parameter plays a significant role in describing the evolution of the assault procedure, which is a crucial assumption to make in the research approach. The Markov model is a modeling approach, which has been widely utilized in several fields, including application performance and reliability research. It is expected that the adversary would select the flaw that gives him or her the best chance of breaching the security goal.

Complete Article List

Search this Journal:
Volume 14: 1 Issue (2024)
Volume 13: 1 Issue (2023)
Volume 12: 4 Issues (2022): 2 Released, 2 Forthcoming
Volume 11: 4 Issues (2021)
Volume 10: 4 Issues (2020)
Volume 9: 4 Issues (2019)
Volume 8: 4 Issues (2018)
Volume 7: 4 Issues (2017)
Volume 6: 4 Issues (2016)
Volume 5: 4 Issues (2015)
Volume 4: 4 Issues (2014)
Volume 3: 4 Issues (2013)
Volume 2: 4 Issues (2012)
Volume 1: 4 Issues (2011)
View Complete Journal Contents Listing