Posted: February 14th, 2023

Cybercrimes have adversely impacted political

Cybercrime
Name:
Institution:

Cybercrime
Cybercrimes have adversely impacted political, economic, and social spaces. As a result, stakeholders in these areas have realized the need to develop effective countermeasures by predicting such crimes and appropriately acting on them as opposed to dealing with the aftermath. According to Brewster et al. (2015), the implementation of cybercrimes takes the use of computers and networks to commit fraudulent and criminal activities. The computers are adopted in the commission of a crime or as a target at the expense of the holder. Consequently, different cybercrimes are directed towards disadvantaging users of the internet, computers, and other devices. The different crimes include phishing scams, invasion of privacy, cyber-stalking, debit or credit fraud, online credential breach, identity theft scams, online harassment, and introduction of malicious software into the network and computer systems. Various cybercrimes are exercised at the expense of users and in the interest of the criminals (Brewster et al., 2015). The overall objective of cybercriminals is to illegally benefit in the misfortune of other people.

According to Johnson (2015), cybercriminals have consistently advanced their skills and prowess in implementing attacks. They manipulate computer systems and networks to retrieve and mine sensitive information that enables them to continuously manipulate the system in their target. Therefore, there is a need to develop effective measures to counter the threat of cybercrime by predicting its occurrence. Hence, they can adopt the appropriate measures to eliminate them in the interest of business and other consumers. The preventive measures should be in the form of regulations and policies that safeguard the computer and network systems (Johnson, 2015). Cybercrimes results in irredeemable losses in money, reputation, and other aspects of life to the government, businesses, and individuals and thus there is a need to effectively predict the occurrence of such events to make it successful in preventing the crimes as opposed to dealing with the aftermath of such crimes.
Predicting Cybercrimes
There is a great need to predict the form and types of cybercrime in government, organization, and individual spaces making it possible to take sufficient measures, regulations and policies counter and eliminate cyber-related crimes. According to Sun et al. (2018), cybercrimes have increasingly become sophisticated making it impossible to find effective tools to deal with the attack thus predicting the crimes the most efficient approaches to eliminate the crimes. Therefore, the government, researchers, and professionals have extensively researched and developed new models and best practices that are more effective in the prediction and prevention of potential cybercrimes. The prediction techniques and approaches to cyber-attacks are regarded as defense strategies. The prediction tools used predict cyber-attacks through the identification of malicious activities and in the computer and network systems (Tavabi et al., 2018). Governments and organizations need to be inclined towards predicting and planning the cyber-attacks that could take place in the future as opposed to depending on the removal and defensive measures.
Attack Graphs
The attack graphs are effective analytical and prediction methods for cybercrimes in government, organization, and individual computer and network systems. Researchers have successfully used attack graphs to predict the probability of cybercrimes. The attack graphs indicate the different ways attackers/ cybercriminals can exploit vulnerabilities to gain illegal access to computers and networks (Zhan, Xu, and Xu, 2015). This data is critically analyzed and evaluated to identify the weaknesses in the systems. The attack graphs present paths that can be used by an intruder into a system to implement successful attacks. The security analysts adopting the attack graphs use algorithms to generate attack graphs that are vital in forensic, defense, and detection operations in cyber-related attacks.
Attack graphs use minimization analysis techniques to enable them to analyze to make decisions on a minimal set of security measures that enhance the safety of the system. More so, the reliability analysis technique is adopted to ensure that analysis can be conducted through simple cost-benefit trade-off based on the probability of the attacks (Singhal and Ou, 2017). The two techniques on the attack graphs ensure that the analyst can effectively compute the probability of intruders initiating success in their attack operations. Attack graph evaluation and analysis is critical in the identification of cybercrime threats and risks that can negatively impact the operations of governments and organizations are detected in advance. This helps to ensure that the appropriate defense strategies are adopted to prevent attack events from taking place.
Capability, Opportunity, and Intent of the attacker (CIO)
The Capability, Opportunity, and Intent of the attacker (CIO) techniques have been increasingly adopted in the computer system and networks to predict cybercrimes events. The CIO technique is vital in identifying the types of attacks that the attacker is likely to use on a system. This fact is reached through the identification of vulnerabilities that the attacker is likely to take advantage of to access the computer and network system (Scherling, 2016). The attackers tend to identify the weakness and loopholes in the system in the interest of capitalizing on them to gain access to the system. Various components of the COI have a different function in predicting and preventing cybercrimes. The capability component predicts services that the attacker is likely to attack based on the previous successful attacks. The opportunity component investigates the chances of the attacker having insider information and the necessary precaution that safeguards the network from such attacks. More so, the intent component evaluates the social influence and motivation of the attacker to decide on the most probable types of attacks that can be adopted (Scherling, 2016). This method has been increasingly used by military and intelligence communities to analyze cyber-related threats thus alerting them in the event of such attacks taking place in their different spaces.
Internet Protocol Addresses (IP)
Internet Protocol Addresses (IP) are vital in ranking the threats and attacks that are prone to different systems thus alerting the likely victims on the approaches to adopt in preventing such attacks. Previously, the IP has been used in ranking shopping and movie sites. The IP ensures government organizations to predict the vulnerabilities in networks as per the malicious source internet protocol addresses (Watters et al., 2012). The IP is a numerical label assigned to every device connected to the network and record the activities taking place in the system. In this regard, activities that pose threats are recorded and ranked to identify the most likely forms of cyber risks and threats. Upon the identification of cyber threats and risks in the system and the globe, the organization can take the relevant counter strategies in the form of techniques and best practices in protecting the network. Internet Protocol Addresses (IP) are adopted by governments to evaluate the cyber operations taking place in the different networks and computers to access (Watters et al., 2012). It also investigates cyber threats occurring in different computers and networks thus inclining government agencies and organizations to take the relevant countermeasures to protect their network systems and computers.

Dynamic Bayesian Network (DBN)
Dynamic Bayesian Network (DBN) techniques are adopted in evaluating cyber risks and threats thus making effective predictions of the threats and vulnerabilities posed to government, organizations, and individuals (Jongsawat, 2016). DBN is a statistical model that estimates over a significant period, thus identifying the patterns of attacks that are likely to be implemented on the systems. Under the DBN different variables are evaluated over a long period in their relation to the computer system and networks. The different variables (types of threat) in the DBN has developed the calculations of internal repressors and previously stated values (Jongsawat, 2016). The complete calculation gives the overall probabilistic demonstration and extrapolation mechanism for the different cybercrimes aspects.
The DBN extends the standard Bayesian networks while incorporating the aspect of time. The DBN network has different features such as Support multivariate time series, log-likelihood, support for time series and sequences, structural learning of temporal models, complex temporal queries, parameter learning of temporal models, post probable sequence, pix temporal and non-temporal variables, and prediction, filtering, smoothing (Jongsawat, 2016). The different features are vital as they are incorporated with variables (forms of cyber threats) and the aspect of time to detect cybercrimes that have a high likelihood of happening in the future. Jongsawat (2016) further explains that the FBI has used the DBN to learns and understand different cybercrimes with time thus they can effectively predict future cybercrimes from the graph and thus adopt the relevant cybercrimes countermeasures.
Attack Strategy Synthesis and Ensemble Predictions of Threats (ASSERT)
The cybercrimes have been effectively predicted using the Attack Strategy Synthesis and Ensemble Predictions of Threats (ASSERT) techniques thus prompting the probable victims to adopt the relevant policies, regulations, and rules to prevent future cybercrimes (Okutan and Yang, 2019). The ASSERT strategy evaluates the observable malicious operations affecting the networks to predict the future occurrence of cyber-related attacks. In this regard, the strategy can create a strategy to differentiate ongoing cybercrimes and respond to the upcoming critical threats before they affect the different organizations. The ASSERT technology used data from different sources in the networks such as system logs and detection system alerts to enhance the recognition and prediction of the existing cyber threats.
ASSERTS techniques use a combination of strategies and techniques such as information theory-based divergence, clustering, and Bayesian learning technologies to develop and refine hypothetical attacks on computer and network systems. Additionally, the data called in the system is fed in the Generative Adversarial Network (GAN) and Long-Short-Term-Memory (LSTM) to analyze the sequential data in the interest of characterizing cyber-attacks (Okutan and Yang, 2019). For instance, the US-based National Science Foundation (NSF) has consistently evaluated the ASSERT in predicting the probability of cybercrimes occurring in different organizations such as US investigative agencies (Chi et al, 2001). Therefore, ASSERTS ensures that cybercrimes data is collected and analyzed to effectively predict the occurrence of crimes in the future thus inclining the affected parties to adopt the necessary countermeasures to safeguard the networks and computer systems.
Cyber Attack Scenario and Network Defence Simulator (CASCADES)
The Cyber Attack Scenario and Network Defence Simulator (CASCADES), enhances the effective prediction of cybercrimes to governments and organizations across the globe. CASCADES presents a modern and advanced way of predicting cybercrimes through the simulation of cyber-attacks scenarios following the renewed criminology theory for cybercriminals (Lever, MacDermott, and Kifayat, 2015). The presented global scenarios enable the technique to adopt the ”what if” analysis thus forecasting the different forms of cyber threats that can occur in the future. CASCADES techniques generate different scenarios through importance sampling methods such as preference and attacker COI as well as the Monte Carlo simulation.
Monte Carlo simulation entails robust technical analysis tools develop exhibit random and chance variables and the same strategy is used in other fields such as engineering and finance. The main importance of sampling includes the fact that the statistical technique used in making estimations for properties of a group of potential outcome probabilities. The simulations are run for different network configurations and multiple attacker types that include random attackers, amateurs, and experts (Chi et al, 2001). The analysis of simulation scenarios is theory and exploratory based that enhances an effective understanding of the scenarios to enable effective prediction of cybercrimes. The NSF has used the CASCADES as a project to evaluate its impacts in the prediction of cybercrimes to enhance the prevention of risk and threat occurrences from taking place (Chi et al, 2001). The CASCADES strategy incorporates different Monte Carlo simulation and importance sampling to identify the cybercrime scenarios thus predicting the probability of cybercrimes from taking place.
Big Data Analytics
Parties prone to cybercrime have adopted sophisticated tools such as big data analytics in detecting and predicting future cybercrimes across the globe. Big data analytic enhances efficient monitoring of aspects of cybercrime thus enhancing their detection and prevention of economic crimes. Cybercrimes are consistently reported and recorded and traced to indicate the frequency and trends across the globe or in different regions (Brewster et al., 2015). The current trend shows that steadiness, declines, and increase of cybercrimes across the globe thus making it possible to detect future cybercrimes. The researchers work differently with the current data on cybercrime-related issues to derive different aspects that enhance prediction of cybercrimes that would decrease or increase based on the technological advancement on the countermeasures as well as the techniques employed by the cybercriminals. For instance, the banking sector in the US has used big data analytics to predict the future aspects of cybercrimes across the globe thus enabling the sector to adopt the necessary countermeasures in form of policies, regulations, laws, and the appropriate tools/techniques (Brewster et al., 2015). Therefore, data analytics employed in predicting future cybercrimes uses the existing data in line with existing technology thus making it effective.
Machine Learning Technology
Additionally, machine learning technology has been advanced to enhance the detection and prediction of cybercrimes before they affect different sectors. The latest machine learning platform, AI2, has been developed by and they are capable of detecting over 85% cyber-attacks. The machine learning platform, AI2, is attained through monitoring the web-scale platform that generates millions of log lines to be used in the prediction of cybercrimes (Prabakaran and Mitra, 2018). The system learns from the previous attacks thus preventing future attacks by alerting the users of the system. The accuracy of the machine learning, in this case, increases with increased attacks on the system and the feedback from the analysts. Human-machine interaction in the evaluation and assessment of cybercrimes enables it to achieve success in the areas that machine learning models have not succeeded. Machine learning operations in detecting crimes are achieved through the reliance of human analysts, increasing the learning capability and incorporation of machine learning algorithms. These facts ensure the machines can evaluate the existing crimes and with incorporations of advanced technology to define the trend of cybercrimes and thus the prediction of the crimes (Prabakaran and Mitra, 2018). For instance, the Global Economic Crime Survey 2016 published by PwC, indicates that organizations in the financial sector have consistently used internal monitoring of their system using machine learning thus effectively predicting future cybercrimes.
Conclusion
There are different tools and techniques used by different parties prone to cybercrimes to predict future cyber-attacks thus adopting the necessary countermeasures thus enhancing the safety of the computer and network systems. The tools and techniques used need to be combined as opposed to using them separately since errors arise from using single techniques. The tools need to be selectively employed in countering attacks to ensure that the tools adopted can effectively operate in different environments as presented by the different fields. The different tools and equipment in the detection and prediction of cybercrimes include attack graphs, CIO techniques, Internet Protocol Addresses (IP), Dynamic Bayesian Network (DBN), Attack Strategy Synthesis and Ensemble Predictions of Threats (ASSERT) techniques, Cyber Attack Scenario and Network Defence Simulator (CASCADES), big data analytics, and latest machine learning platform, AI2. The different techniques and tools used in the prediction of cyber-crimes have different strengths and weaknesses and thus they need to be effectively used and used to capitalize on their strength while reducing their weaknesses. This approach ensures that cybercrimes are predicted and detected before they occur thus making the right decisions to prevent them. The prevention through the detection of crimes is far much better as opposed to dealing with attacks that have already taken place. The prediction and prevention save the concerned parties from losses that can result from the attacks.

References
Brewster, B., Kemp, B., Galehbakhtiari, S., & Akhgar, B. (2015). Cybercrime: attack motivations and implications for big data and national security. In Application of Big Data for National Security (pp. 108-127). Butterworth-Heinemann.
Chi, S. D., Park, J. S., Jung, K. C., & Lee, J. S. (2001, July). Network security modeling and cyber attack simulation methodology. In Australasian Conference on Information Security and Privacy (pp. 320-333). Springer, Berlin, Heidelberg.
Johnson, T. A. (Ed.). (2015). Cybersecurity: Protecting critical infrastructures from cyber attack and cyber warfare. CRC Press.
Jongsawat, N. (2016). Dynamic Bayesian Networks for Information Security. Engineering Journal of Siam University, 17(1), 40-51.
Lever, K. E., MacDermott, Á., & Kifayat, K. (2015, December). Evaluating interdependencies and cascading failures using distributed attack graph generation methods for critical infrastructure defence. In 2015 International Conference on Developments of E-Systems Engineering (DeSE) (pp. 47-52). IEEE.
Okutan, A., & Yang, S. J. (2019). ASSERT: attack synthesis and separation with entropy redistribution towards predictive cyber defense. Cybersecurity, 2(1), 15.
Prabakaran, S., & Mitra, S. (2018, April). Survey of analysis of crime detection techniques using data mining and machine learning. In Journal of Physics: Conference Series (Vol. 1000, No. 1, p. 012046). IOP Publishing.
Scherling, M. (2016). Practical risk management for the CIO. CRC Press.
Singhal, A., & Ou, X. (2017). Security risk analysis of enterprise networks using probabilistic attack graphs. In Network Security Metrics (pp. 53-73). Springer, Cham.
Sun, N., Zhang, J., Rimba, P., Gao, S., Zhang, L. Y., & Xiang, Y. (2018). Data-driven cybersecurity incident prediction: A survey. IEEE Communications Surveys & Tutorials, 21(2), 1744-1772.
Tavabi, N., Goyal, P., Almukaynizi, M., Shakarian, P., & Lerman, K. (2018, April). Darkembed: Exploit prediction with neural language models. In Thirty-Second AAAI Conference on Artificial Intelligence.
Watters, P. A., McCombie, S., Layton, R., & Pieprzyk, J. (2012). Characterising and predicting cyber attacks using the Cyber Attacker Model Profile (CAMP). Journal of Money Laundering Control.
Zhan, Z., Xu, M., & Xu, S. (2015). Predicting cyber attack rates with extreme values. IEEE Transactions on Information Forensics and Security, 10(8), 1666-1677.