2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

Augmenting Cybersecurity and Fraud Detection

Using Artificial Intelligence Advancements
Keshaw Kumar Kuldeep
Department of Computer Science and Engineering Department of Computer Science and Engineering
Sharda University Sharda University
Greater Noida, India Greater Noida, India
2023 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS) | 979-8-3503-0611-8/23/$31.00 ©2023 IEEE | DOI: 10.1109/ICCCIS60361.2023.10425069

[email protected] [email protected]

Bharat Bhushan Kuldeep

Department of Computer Science and Engineering Department of Computer Science and Engineering
Sharda University Sharda University
Greater Noida, India Greater Noida, India
[email protected] [email protected]

Abstract—In 2023, the rapid advancement of Generative AI various aspects of our everyday lives. Because AI relies on
(GenAI) models, including ChatGPT and Google Bard, has data, it's unavoidable that AI systems will need to access and
reshaped the digital landscape. GenAI tools are now used both handle our private information in many everyday
defensively and offensively, raising crucial questions about their applications. For example, when people want financial help
societal, ethical, and privacy implications. Artificial Intelligence like loans, they have to share their personal and money
(AI) plays a pivotal role in the Fourth Industrial Revolution, details. Also, when you visit a website, it might collect data
enhancing the security of computer networks. This study about your online activities to show you personalized ads.
investigates these connections, finding that e-Governance This increase in using personal data has made people worry
partially mediates the link between AI and cybersecurity, with
about their privacy because the data can be sensitive. Even if
stakeholders influencing this dynamic. AI's increasing
integration into daily life, spanning healthcare, finance, and
it doesn't directly reveal who you are, AI can still find out
surveillance, emphasizes the importance of data privacy. The sensitive info from your data. To solve this, one way is to
handling of extensive datasets by AI systems brings about make the records anonymous, so they don't directly link to
unique privacy and ethical challenges, necessitating stringent individuals [1]. On the other hand, privacy by policy involves
safeguards. This concise abstract underscore the profound setting rules to protect privacy. This includes telling users
impact of GenAI models on cybersecurity in 2023, emphasizing their data will be used and making sure everyone follows
their implications for society, ethics, and privacy. It highlights privacy laws.
nuanced relationships between AI, e-Governance, and
Cybersecurity is now very important. It's all about
cybersecurity, while acknowledging the need for robust privacy
protection. The paper calls attention to the potential of
keeping computer networks safe from possible dangers in
advanced techniques to fortify cybersecurity against ever- today's modern world. A cyber-attack is when someone
evolving threats. intentionally tries to harm computer networks, data,
programs, and electronic information [2]. This can lead to
Keywords—Artificial Intelligence , Cyber-security,GenAI groups or entities promoting violence against innocent
people. As technology advances, so do cyber threats. This
I. INTRODUCTION means the need to create new ways to stop them. Cyber-
Artificial Intelligence (AI) and Machine Learning (ML) attacks are said to be happening more often in industries,
have changed The computers work in the past ten years. causing damage to important systems and costing a lot of
They've made big progress, like learning on their own and money [3]. AI applications can have a good impact on the
doing really smart things, not just following orders. The cyber abilities and security of countries, regions, and non-
newest advance in AI technology is Generative AI. It uses government groups. Smart cities offer creative solutions to
deep neural networks to learn from a lot of data and create various problems that city administrations encounter.
new content that follows the same patterns and structures . However, introducing Information and Communication
Right now, the tech industry is trying to build really advanced Technology into the city's infrastructure can bring about risks
Large Language Models (LLMs) that can have conversations and challenges. People often use unsecured Wi-Fi networks
just like humans. Artificial intelligence (AI) means creating for activities like checking email or online banking, making
computer programs that can do things like understand speech, them vulnerable to cybercrimes like hacking, service
see things, and solve problems, similar to The humans do. disruptions, and data breaches. Cybersecurity technologies
Machine learning is a part of the artificial intelligence where are crucial for safeguarding e-Government services, and they
computers learn by themselves from data and don't need step- are a key factor in classifying cities as safe in the global
by-step instructions. Once these strong models learn context [4]. Recently, cybersecurity experts have been
successfully, they can make smart choices and enhance looking into Artificial Intelligence (AI) can enhance

ISBN:979-8-3503-0611-8/23/$31.00 ©2023 IEEE

1207
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.
 2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

cybersecurity. At the same time, cybercriminals are using AI geometry, and a series of blocks-world issues were created
to carry out more advanced cyberattacks while covering their by early AI applications.
tracks. AI-based cybersecurity solutions have the potential to
better defend against these attackers and reduce the chances
of data breaches or even prevent them. AI has come a long A. Background
way since it first appeared in the 1950s, with notable
Agent-based artificial intelligence (AI) or "bots"—
developments like machine learning and deep learning.
software that acts like humans—became increasingly popular
Nowadays, AI is used in a wide domain of fields such as
in late 1990s during the rise of Internet. Search engines,
agriculture, space exploration, law, and manufacturing,
online directories, and recommendation services have all
leading to exciting research findings and practical systems
benefited from the creation of ethical bots that crawl the
[5].
Internet. They offer vandalism defense for Wikipedia articles
The rapid advancement of Artificial Intelligence (AI) and where anyone can participate as an author. In contrast to,
Machine Learning (ML) impacted the privacy and security of malevolent bots have also surfaced to publish spam and
personal data in various everyday applications. The use of AI transmit malware, cheat at online games, and more. In order
in everyday applications affect individuals' privacy, reverse-engineer the game code while simulating online
particularly in contexts such as financial transactions and games, bot programmers examined the traffic flow between
online activities. The potential risks associated with AI's the game console and server[6]. Instead of continuously
ability to extract sensitive information from anonymized sending messages, the bots that posted spam imitated human
data, and how can these risks be mitigated. The emerging online activity by browsing the pages before posting in
cyber threats in the modern world, and how are they evolving forums.
with technological advancements. AI technologies be
B. AI Development Life Cycle
leveraged to enhance cybersecurity, detect and prevent
cyberattacks, and safeguard the data and systems of To solve sophisticated problems in the data driven
governments and organizations. The specific challenges and approach and without the need for explicit humans
vulnerabilities arise in the context of smart cities and their programming is the objectives of an AI system. For example,
information and communication technology infrastructure, it helps radiologists decrease their diagnosis time. AI system
and how can AI solutions contribute to addressing these learned from the data to predictions. According to historical
challenges. These research questions encompass the data, other types of AI systems can incorporate real time data
overarching issue of how AI impacts privacy and make decisions for human input is trained by ML algorithms.
cybersecurity, while the specific questions delve into the Different stages of AI lifecycle is data collection, pre-
nuances of privacy concerns and cybersecurity challenges in procession, analyze. In this context, the development of AI
the context of AI and modern technology. The research paper systems is collection of information of individuals of data
can then proceed to answer these questions by presenting subjects. The purpose and means of personal data processing
relevant findings, analyses, and insights. is determined by data controller which is defined by the
GDPR. Process and possess personal data on the behalf of the
The paper structure are as the follow: Section 2 presents the controller for data processor may also authorized by the data
artificial intelligence and the life cycle of the artificial controller.
intelligence, section 3 elaborate the cybersecurity and the
threat foundation and the challenges in smart cities and how
it is playing crucial role in the cybersecurity, section 4
elaborate the ai for cybersecurity followed by section 4.1
impact of the cybersecurity and privacy and applying
artificial intelligence to cybersecurity for different
applications domain.
II. ARTIFICIAL INTELLIGENCE
AI is interested in machines can act or think appropriately
from a given set of data. This all-encompassing definition
takes into account how closely machine can behave or think
like human. The Turing Test, at the other extreme of
spectrum, establish benchmark for AI. According to this
standard, a computer that communicates with a person is
regarded to be intelligent when the person is unable to
identify whether responseare coming from a human or
computer.AI encompasses compute disciplines such game
theory, mathematics, machine learning, automated reasoning,
logic, knowledge representation and natural language
processing at both ends of the computing spectrum. Thinking
. AI Development Life Cycle
computers that could solve checker games, puzzles like

1208
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.
 2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

1) Project Planning: Project Planning is necessary for learn from data during the training phase while optimizing an
the AI system for outline the objectives in prior to any model objective function. Typically, AI models comprise three
development. The AI will address for identify the relevant components: input, output, and the model itself. The output
use case and defining the scope of system is included by this. generated from the model id specifically categorized the
Therefore, the fundamental step of AI life cycle is decided by model which is known as classification model. The model is
the required data and understanding the project objectives. said to be regression model if the output is a continuous
The project objectives for gather comprehensive information value. The extraction of the relationship among input and
and other development details is necessary for developing an output help the model to learn efficiently and perform better.
efficient AI application, such as potential system users and For author/s of only one affiliation (Heading 3): To change
data sources. Failure to specify the project's goals will the default, adjust the template as follows.
inevitably cause delays in execution and may have a 5) Model Deployment: The outcome of the developed
detrimental effect on performance. A greater scope of product is highly related on the satisfaction of data subjects
diagnosing various ailments using magnetic resonance and stakeholders. Consequently, the ultimate system should
imaging scans, such as spinal cord and brain injuries, may not align with the initial phase's intended objectives. Conducting
be effective in the case of medical image diagnosis, for thorough testing before deployment can help guarantee the
example. According to the data that is available, training the system's precision and the model's impartiality. The end
AI model to recognize a particular injury may produce results users' ease of use should be taken into account for successful
that are more accurate. deployment. For instance, a suitable application that enables
2) Data Collection: After the identification of the users to interact with the finished model naturally can be
objectives and definition of the scope of the project data created. Real-time use of the AI model may be necessary in
collection is done. Here the quality of the data is an important specific circumstances, such as when identifying traffic
factor because the whole AI and ML system are data driven. offenses, and such demands must also be addressed at these
The process of data collection helps the model to understand days. Additionally, model deployment should guarantee that
to implementation of the idea and to get the required the AI algorithm has access to the necessary memory and
information for data scientist to build AI model. Now, if talk computational resources. Complex models, for instance, may
about the generalization of the AI models and biasness of the not be appropriate for deployment on constrained hardware
AI to ensure that the dataset is collective, it should contain devices, such as smart phones are better for deployment on
diverse representation of all the related fields[7]. There are the cloud.
several types of data collective data such as catagorial, time
series, and text, each data type is different in many respects III. CYBERSECURITY
they may differ on the basis of storage , maintenance Cybersecurity is like a digital shield that protects your
,processing ,it may also have the risk of vulnerability. To computer and personal information from bad guys on the
avoid these serious issues data storage should follow the internet. It's all about using special tools and strategies to
standard of secured procedure. make sure your data and online activities stay safe and secure.
Just like you lock your front door to keep out intruders,
3) Data Preparation:The procedure followed in
cybersecurity helps keep your online world safe from hackers
collection of the data or the preparation of the data can affect and other digital threats.
the the training time as well as performance of the model. The
performance of the system can be enhanced by selecting the A. Threat Foundation
right data type, it also depend on object of a project and the In 2021 there were damages around $6 trillion all around
requirement of the model. Preprocessing steps such as feature the globe in the area of Cybercrime. If we treat this damage
extraction, data cleaning feature selection. All the steps as a country GDP, it would have the world's third largest
discussed above helps the data scientist to cope up with the economy after the United States and China. As per Cyber
incomplete data which misses some values, out liners etc. the Security Experts, global crime in field of Cybercrime
increases 15% per year since the last haul of five years. There
visualied data worl in the favour of the data scientist as to
is a prediction that by 2025, the economy of the Cybercrime
help us to detect and remove the unwanted and wrong sample. sector increases approximately up to $10.5 trillion, which is
Data integration is one of the important stage of formation of actually $3 trillion in 2015[8].
AI model, by this user is provided with standard data format.
It require sematic interoperability which means the ability of Cyber Security is actually a process which deals in
defending ICT systems from various Cyber threats. Cyber-
system use and exchange of data in a platform and make
Attack is the activity of preying someone’s electronic
system capable to use the data from different sources. information without the consent of that particular individual
4) Model Development: Choosing the right model and or organization. Preying on networks, electronic information
approach for system development is crucial in achieving systems, or infrastructure is illegal. By this process one can
optimal performance. Some algorithms, like k-nearest Stole the data or information, can alter the information and
neighbour, rely on extracted rules from samples and do not can even destroy it. There is primarily one difficulty, which
necessitate a training phase. In contrast, AI-based models

1209
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.
 2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

is the understanding of various cyber-attacks and the IV. AI FOR CYBERSECURITY

strengthening of supplementary defensive methods.
A. Impact of GenAI in Cybersecurity and Privacy
There are mainly three security variables in account to The generalization power in AI replacing the traditional
take any attack: threats, the one who is the attacking approaches with new technology. The upgrading technology
vulnerabilities, or the loop holes they are trying to attack and also elevating the cyber thereat actor’s sophistication.
impacts, or what can be the possible damage or assault it can Intrusion attempts, cyberspace faced relatively are the
make. An incident of the Cyber Security threatens the traditionally attempts to break into the system. The Cyber-
integrity, confidentiality, or availability of data or attacks is more powerful than every as AI/ML has upgrade
information assets and systems. There are several ways that a effectively. The GenAI has gained interest in cyber offense
computer system or any mobile device's security might be and cyber defense. The evolving of GenAI is benefiting both
violated, including through data deletion, unauthorized attacker and defenders in cybersecurity. ChatGPT can be
access, and information manipulation. Threats include used to safeguard the systems from intruders. These tools are
information on all of the aforementioned security violations, trained on massive amount of data that includes attack
potential dangers and hazards, and attacks, which explain patterns, cyber threat data and vulnerabilities data. Cyber
attempts to commit a Cyber-crime or violation. defenders enhance their intelligence by identifying emerging
There are some precautions which can be take in order to threats and extracting data. The GenAI can be used to analyze
maintain the information and communication technology the network traffic data, system output or large volume of log
secure, including the raw data and information it contains as files. This helps defenders to automate and speed up the
well as the processing and transmission linked to the virtual incident response. GenAI are helpful in generating the secure
(software) and actual (hardware) components of the system. code and written secure code. Attackers use the GenAI to
The degree which tells about security of system is associated create different malicious code, phishing attack and social
in the field of Cybersecurity under professional endeavor. engineering attack.
B. Challenges in smart cities B. Applying Artificial intelligence to cybersecurity for
Most of the smart cities are the captivating concept which different applications domain
is characterized by its intelligent features. The scope of The number of users of Internet has been increased in the
improvement has grown beyond cost reduction, resource size and the number of applications is also increased. Today
usage, and increased urban economic efficiency. All the Internet is important utility similar to the gas, water and
components of the city are integrated by intelligent gadget electricity. The increasing risk of exposed of cyberattacks as
and digital technologies or communication and information more devices connect to Internet. The AI play major role in
technology to better the delivery of different services. As the the three domains of the artificial intelligence which are
conventional urban villages or cities is being transformed into critical infrastructure, Internet of things (IOT) and the
advanced or smart cities, the standard of the citizens has Internet. AI perspective see the harmful pattern are different
upgraded[9]. from the legitimate traffic which are commonly cyberattacks.
To differentiate between legitimate traffic and malicious
For a government to be smart there are some impotent traffic intrusion detection develop by artificial intelligence
aspects to be considered such as smart offices, smart services technique because they are capable of read large amount of
and the most important decision making. These aspects data. Cyberattacks targets are users, business logic and
enhance the overall performance of governance, and establish network infrastructure. In DoS attacks attackers prevent the
a bilateral collaboration in government and citizens which access of service from serving client requests. In botnet attack
eventually upgrade the standard of living for citizens. Various the attacker compromise hosts and request to execute tasks
electronic data or information and online services are being on that machine and then controls it. In a DoS attack, hacked
offered by smart public services which result in the enhanced workstations were utilized to overload with a high number of
standard of living and satisfaction of the public. The progress requests, leaving no resources to handle legitimate users'
of the smart economy can result in development of smart queries. DoS attacks are becoming more dangerous as they
cities, increase in efficiency of economics of urban cities, and become more complicated and execute on many PCs, IoT
create sustainable employment opportunities. Smart health devices, and mobile devices. DoS attacks can be avoided
care has revolutionized the way the living, there are some using the proposed detection mechanism. The New Denial of
useful technologies that measure the step walked in a day services techniques attack new services. The new DoS attack
which can prevent us from serious cardiac disease. Smart can be used on Smart meters [10]. DoS attack can be detect
education has made the life smoother ever, now anyone can using deep learning methods.
learn any skill by just sitting in the couch using high quality
animation videos, which saves time and also reduces stress. V. RECENT ADVANCES
Smart transport system which are multifaceted and managed
Aditi et al. [11] suggested to the data mining inheritalble
digitally enhance the development of urban cities and
and programing are used to identify the fraud. Kapadiya et al.
decision making. Smart security offers a variety of health
[12] proposed AI and blockchain based secure intelligent
advantages, including detection, alert, emergency assistance,
system for detecting the fraud. HJ et al. [13] suggested the
and many other features that are very effective in protecting
classifications of telecom fraud and its impact. Ahmed et al.
people and maintaining security.
[14] presents algorithms like XGBoost , KNN , Naive Bayes,

1210
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.
 2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

Logistic regression and random forest to correctely identifies various sectors, it has raised pertinent concerns about data
fraudulent. Mitra et al. [15] implemented the hybrid QNN and privacy and ethical data processing.
TDA detection for the credit card fraud detection. Sethi et
al.[16] proposed mangagment procedure and medical record Additionally, the integration of Explainable Artificial
storage , audit abilty , data security immutablity and Intelligence (XAI) into the field of cybersecurity introduces
traceabilty to detect fraud in medical insurance. Bhowmik et both benefits and challenges, as it enhances transparency
al. [17] proposed a deep belief network , nonlinear embedded while potentially exposing vulnerabilities to adversarial
clustering to detect farud transactions. Wu et al. [18] attacks. As cybersecurity threats become more sophisticated,
classified fraud detection model into two category one is traditional defenses are increasingly inadequate,
adversarial and another is unsupersvised. Dhieb et al.[19] necessitating innovative approaches. Machine learning and
proposed the development of the XGBoost and deep learning, bolstered by cryptographic techniques, hold
aforementioned insurance and machine learning algorithms promise for addressing the ever-evolving cyber threats.
to boots the the performance of the algorithms. Capuano et This research underscores the dynamic and complex
al.[20] identified the different type of fraud which can take landscape of AI, privacy, and cybersecurity. The findings
place in order to steal money from people. Wagle et al. [21] emphasize the need for ongoing research and proactive
proposed examines the credibility of the platform using an measures to strike a balance between the advantages and risks
explainable AI-assisted Multimodal Credibility Assessment posed by AI and ML technologies. It is imperative to
System where the blog is hosted. Pastor et al. [22] proposed prioritize transparency, stakeholder involvement, and ethical
a technology solution for Ads.chain to lack of transparency considerations to navigate this transformative digital era
problem in programmatic advertising. Gu et al. [23] safely. As the world embraces the potential of AI and ML,
discussed the effects, capabilities, and scenarios of AI these insights will inform the development of resilient
algorithms used in academic fraud. Zhou et al. [24] proposed cybersecurity strategies and the safeguarding of individuals'
the user-centred explanations financial fraud detection based privacy in the face of emerging technological challenges.
on explainable artificial intelligence methods. Momeny et al.
[25] proposed regular CNN-based methods with compared to REFERENCES
LAII-v4 CNN classifiers. Cite the table in paragraphs [1] M. Ruckenstein and J. Granroth, ‘‘Algorithms, advertising and the
intimacy of surveillance,’’ J. Cult. Econ., vol. 13, no. 1, pp. 12–24, Jan.
TABLE I. SUMMARY OF RECENT ADVANCES 2020.
[2] B. Alhayani, H. J. Mohammed, I. Z. Chaloob, and J. S. Ahmed,
‘‘Effectiveness of artificial intelligence techniques against cyber
Author year description
security risksapply of IT industry,’’ Mater. Today, Proc., vol. 531, pp.
Aditi et al. [11] 2022 Data mining, Inheritable, fraud 1–6, 2021, doi:10.1016/j.matpr.2021.02.531
Kapadiya et al. [12] 2022 Ai, blockchain, secure intelligent
system [3] Kumar A., Bhushan B., Malik A., Kumar R. (2022) Protocols,
Solutions, and Testbeds for Cyber-Attack Prevention in Industrial
HJ et al. [13] 2022 Telecom fraud, its classification
SCADA Systems. In: Pattnaik P.K., Kumar R., Pal S. (eds) Internet of
Ahmed et al. [14] 2023 XGBoost, KNN, Naïve bayes Things and Analytics for Agriculture, Volume 3. Studies in Big Data,
Mitra et al. [15] 2021 Hybrid QNN, TDA detection vol 99. Springer, Singapore. https://doi.org/10.1007/978-981-16-6210-
Sethi et al.[16] 2022 Data security, traceability 2_17
Bhowmik et al. [17] 2022 Deep belief network, nonlinear
[4] Kumar A., Bhushan B., Nand P. (2022) Preventing and Detecting
embedded clustering
Intrusion of Cyberattacks in Smart Grid by Integrating Blockchain. In:
Wu et al. [18] 2021 Fraud detection model Sharma D.K., Peng SL., Sharma R., Zaitsev D.A. (eds) Micro-
Dhieb et al.[19] 2020 Machine learning algorithms, Electronics and Telecommunication Engineering. Lecture Notes in
Capuano et al.[20] 2022 Types of fraud Networks and Systems, vol 373. Springer, Singapore.
Wagle et al. [21] 2021 Explainable ai-assisted multimodal https://doi.org/10.1007/978-981-16-8721-1_12
credibility assessment system [5] D. Yuhas. (Oct. 2017). Doctors Have Trouble Diagnosing
Pastor et al. [22] 2021 Transparency problem in Alzheimer’s. AI Doesn’t. NBC News. Accessed: Dec. 25, 2019.
programmatic advertising [Online].Available: https://www.nbcnews.com/mach/science/doctors-
Gu et al. [23] 2022 Capabilities of AI algorithms havetrouble-diagnosing-alzheimer-s-aidoesn-t-ncna815561
Zhou et al. [24] 2023 Financial fraud detection [6] A. R. Kang, J. Woo, J. Park, and H. K. Kim, ‘‘Online game bot
Momeny et al. [25] 2022 CNN-based methods detection based on party-play log analysis,’’ Comput. Math. Appl., vol.
65, no. 9, pp. 1384–1395, May 2013
[7] R. Schwartz, L. Down, A. Jonas, and E. Tabassi, ‘‘A proposal for
VI. CONCLUSION identifying and managing bias within artificial intelligence,’’ Inf.
In conclusion, the rapid evolution of Generative AI Technol. Lab., Nat. Inst. Standards Technol., Gaithersburg, MD, USA,
2021, pp. 1–24.
(GenAI) models, such as ChatGPT and Google Bard, has
[8] ] S. Morgan. (2020). Special report: Cyberwarfare in the C-suite, 1666
been a pivotal development in the digital landscape. These online. Cybercrime Magazine. [Online]. Available:
advancements have significant implications for https://cybersecurity 1667 ventures.com/cybercrime-damages-6-
cybersecurity, privacy, and ethical considerations. The trillion-by-2021/
multifaceted relationship between AI, e-Governance, and [9] Mehta S., Bhushan B., Kumar R. (2022) Machine Learning
cybersecurity reveals that stakeholders play a critical role in Approaches for Smart City Applications: Emergence, Challenges and
shaping the security of cyberspace. Moreover, while Opportunities. In: Balas V.E., Solanki V.K., Kumar R. (eds) Recent
Advances in Internet of Things and Machine Learning. Intelligent
Artificial Intelligence has found widespread application in Systems Reference Library, vol 215. Springer, Cham.
https://doi.org/10.1007/978-3-030-90119-6_12

1211
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.
 2023International Conference on Computing, Communication, and Intelligent Systems(ICCCIS)

[10] Bhowmik T., Bhadwaj A., Kumar A., Bhushan B. (2022) Machine [25] Momeny, M., Neshat, A. A., Jahanbakhshi, A., Mahmoudi, M.,
Learning and Deep Learning Models for Privacy Management and Ampatzidis, Y., & Radeva, P. (2023). Grading and fraud detection
Data Analysis in Smart Cites. In: Balas V.E., Solanki V.K., Kumar R. of saffron via learning-to-augment incorporated inception-V4 CNN.
(eds) Recent Advances in Internet of Things and Machine Learning. Food Control, 147, 109554.
Intelligent Systems Reference Library, vol 215. Springer, Cham. https://doi.org/10.1016/j.foodcont.2022.109554.
https://doi.org/10.1007/978-3-030-90119-6_13
[11] A. Aditi, A. Dubey, A. Mathur and P. Garg, "Credit Card Fraud
Detection Using Advanced Machine Learning Techniques," 2022 Fifth
International Conference on Computational Intelligence and
Communication Technologies (CCICT), Sonepat, India, 2022, pp. 56-
60, doi: 10.1109/CCiCT56684.2022.00022.
[12] K. Kapadiya et al., "Blockchain and AI-Empowered Healthcare
Insurance Fraud Detection: an Analysis, Architecture, and Future
Prospects," in IEEE Access, vol. 10, pp. 79606-79627, 2022, doi:
10.1109/ACCESS.2022.3194569. R. H J and Mohana, "Fraud
Detection and Management for Telecommunication Systems using
Artificial Intelligence (AI)," 2022 3rd International Conference on
Smart Electronics and Communication (ICOSEC), Trichy, India, 2022,
pp. 1016-1022, doi: 10.1109/ICOSEC54921.2022.9951889.
[13] R. H J and Mohana, "Fraud Detection and Management for
Telecommunication Systems using Artificial Intelligence (AI)," 2022
3rd International Conference on Smart Electronics and
Communication (ICOSEC), Trichy, India, 2022, pp. 1016-1022, doi:
10.1109/ICOSEC54921.2022.9951889.
[14] A. N. Ahmed and R. Saini, "Detection of Credit Card Fraudulent
Transactions Utilizing Machine Learning Algorithms," 2023 2nd
International Conference for Innovation in Technology (INOCON),
Bangalore, India, 2023, pp. 1-5, doi:
10.1109/INOCON57975.2023.10101137.
[15] S. Mitra and K. Rao JV, "Experiments on Fraud Detection use case
with QML and TDA Mapper," 2021 IEEE International Conference on
Quantum Computing and Engineering (QCE), Broomfield, CO, USA,
2021, pp. 471-472, doi: 10.1109/QCE52317.2021.00083.
[16] B. K. Sethi, P. K. Sarangi and A. S. Aashrith, "Medical Insurance Fraud
Detection Based on Block Chain and Machine Learning Approach,"
2022 Fourth International Conference on Emerging Research in
Electronics, Computer Science and Technology (ICERECT), Mandya,
India, 2022, pp. 1-4, doi: 10.1109/ICERECT56837.2022.10060811.
[17] A. Bhowmik, M. Sannigrahi, D. Chowdhury, A. D. Dwivedi and R.
Rao Mukkamala, "DBNex: Deep Belief Network and Explainable AI
based Financial Fraud Detection," 2022 IEEE International Conference
on Big Data (Big Data), Osaka, Japan, 2022, pp. 3033-3042, doi:
10.1109/BigData55660.2022.10020494.
[18] T. -Y. Wu and Y. -T. Wang, "Locally Interpretable One-Class
Anomaly Detection for Credit Card Fraud Detection," 2021
International Conference on Technologies and Applications of
Artificial Intelligence (TAAI), Taichung, Taiwan, 2021, pp. 25-30, doi:
10.1109/TAAI54685.2021.00014.
[19] N. Dhieb, H. Ghazzai, H. Besbes and Y. Massoud, "A Secure AI-
Driven Architecture for Automated Insurance Systems: Fraud
Detection and Risk Measurement," in IEEE Access, vol. 8, pp. 58546-
58558, 2020, doi: 10.1109/ACCESS.2020.2983300.
[20] N. Capuano, G. Fenza, V. Loia and C. Stanzione, "Explainable
Artificial Intelligence in CyberSecurity: A Survey," in IEEE Access,
vol. 10, pp. 93575-93600, 2022, doi: 10.1109/ACCESS.2022.3204171.
[21] V. Wagle, K. Kaur, P. Kamat, S. Patil and K. Kotecha, "Explainable
AI for Multimodal Credibility Analysis: Case Study of Online Beauty
Health (Mis)-Information," in IEEE Access, vol. 9, pp. 127985-
128022, 2021, doi: 10.1109/ACCESS.2021.3111527.
[22] A. Pastor, R. Cuevas, Á. Cuevas and A. Azcorra, "Establishing Trust
in Online Advertising With Signed Transactions," in IEEE Access, vol.
9, pp. 2401-2414, 2021, doi: 10.1109/ACCESS.2020.3047343.
[23] Gu, J., Wang, X., Li, C., Zhao, J., Fu, W., Liang, G., & Qiu, J. (2022).
AI-enabled image fraud in scientific publications. Patterns, 3(7),
100511. https://doi.org/10.1016/j.patter.2022.100511.
[24] Zhou, Y., Li, H., Xiao, Z., & Qiu, J. (2023). A user-centered
explainable artificial intelligence approach for financial fraud
detection. Finance Research Letters, 58, 104309.
https://doi.org/10.1016/j.frl.2023.104309.

1212
Authorized licensed use limited to: University of South Australia. Downloaded on May 15,2024 at 03:14:15 UTC from IEEE Xplore. Restrictions apply.