DEVELOPMENT OF AN INTELLIGENT CHATBOT FOR CYBERSECURITY

Akintola G.K.1, Olatunde Y. O.,2,3, Obiwusi K.Y.2, Afolabi G. K.2, Abdulkarim Z. A.2, Tewogbade
A. A.4, Omotosho L. O.4 & Oladipo-Tanimowo A. A.3
1 Department of Computer Science, University of Ilorin, Ilorin, Kwara State, Nigeria.
2 Department of Computer Science, Summit University Offa, Offa, Kwara State, Nigeria.
3 Decpartment of Cyber Security, Osun State University, Osogbo, Osun State, Nigeria.
4 Decpartment of Computer Science, Osun State University, Osogbo, Osun State, Nigeria.

Corresponding Author: Olatunde Y. O.; [email protected] &

[email protected]

Abstract

The development of an intelligent chatbot to enhance knowledge is certainly an innovative

approach in the field of educational technology. The aim of this project is devising a chatbot that
applies NLP and ML for personalized and interactive educational materials. The chatbot is
intended to provide assistance in real-time, answer questions and give explanations of complex
cyber security topics on the way. The chatbot will be trained to cover major aspects of cyber
security including threat detection, risk management, ethical hacking, and data protection just to
name a few. The chatbot can mimic real-life situations and apply adaptive learning algorithms
for evaluating the user's understanding and adjust the level of the topics and questions based on
those results. This move does not only condense knowledge but also diagnose understanding and
correct them. The intelligent chatbot makes a teacher’s job more efficient as it gives the students
a practical, hands-on experience that may sometimes be lacking in the traditional classroom
setting. On one hand though, its availability all the time means that students can study
uninterrupted. In contrast to that, the development of the digital project will also focus on issues
as data accuracy, user data protection, and keeping users engaged. Nevertheless, the integration
of the chatbot to cyber security education is usually the most effective way to make sure that the
learning is as interactive and individual-oriented as possible.

Keywords: Artificial Intelligent, Cyber Security

1.0 INTRODUCTION
The integration of technology into our daily lives has brought about numerous benefits,

but it has also exposed individuals, businesses, and critical infrastructures to a growing

array of cyber threats (Mohamed, 2023). Despite efforts to fortify digital defenses, users

often struggle to keep up with the dynamic cyber threat landscape, leading to a

significant gap in cyber security awareness. Traditional cyber security education and

training methods are often inadequate to keep pace with the evolving tactics of

cybercriminals.

Recent advancements in artificial intelligence (AI) and natural language processing

(NLP) have opened new avenues for addressing this cyber security awareness gap.

Chabot’s like "Chabot Sec" (CBS) and the "SecBot" prototype have emerged as promising

tools for providing real-time information security advice and engaging users in cyber

security management(Bulin Shaqiri, 2021). To respond to customer inquiries precisely

and quickly, these chatbot employ knowledge-based searches and keyword extraction

strategies.

Furthermore, (Al-Hawawreh et al., 2023)explore the potential for sophisticated AI

systems to support security analysts in analyzing, planning, and countering cyber threats

by deploying models like ChatGPT for cybersecurity applications. Nonetheless, there are

still difficulties, such as the requirement for data usage transparency, problems with
correlational power in specific tools, such as Elastic search (Michel et al. 2023), and the

difficulty of handling various content domains (Hamad & Yeferny, 2020)

Hackers are increasing daily therefore users need to also take precaution on how to

prevent them self to avoid cyber-attack or an intruder from getting access to their

information or properties. Technology an advance and hackers are working hard to get

users properties or system but with the use of artificial intelligence, scientist are working

hard to prevent users from intruders. Therefore the integration of cyber security as an

educator for user will be a greater way to prevent user on any form of attack, it will

educate user on the precaution to take to avoid any form of attack or how to resolve any

kind of cyber related issues.Traditional methods of monitoring and safeguarding against

these harmful behaviours are ineffective. Wiafe et al., (2020) identified deficiencies in

current cybersecurity methods to combat increasingly sophisticated cybercrimes.

As a result, criminal activities like cybercrime and unlawful business operations thrive

on the Dark Web because it is less monitored. Using data collected from four dark web

discussion boards identified as having terrorist-related content, Saini & Bansal, (2019)

demonstrate how terrorist groups launder their dark web activity through private social

media. Exposed personal information is traded and exchanged in underground forums.

Forums have played a role in providing information on data breaches. Fang et al., (2019)

state that the trend is improving in reporting results of data theft through posts in the

forums.
The primary focus of this study is on the application of artificial intelligence to monitor

the Dark Web for cybercrime and other illegal activities. Section 1.2 presents how the dark

web works. Section 1.3 explain how to access the dark web and section 1.4 discussed the

TOR hidden services. In section 1.5, existing dark web monitoring approaches discussed

and 1.6 focused on the application of AI in Dark Web. Section 2.0 give extensive review

of related work and section 3.0 explain the methodology. Crawling and data acquisition,

Data pre-processing, AI model building and analysis of the dark web is presented in

section 3.1, 3.2, 3.2 and 3.4 respectively. Section 4.0 presents the result and discussion.

The Model Training and Evaluation is explained in section 4.1 while section 4.2 covers

the development and experiment interface. Section 5.0 discussed the summary and

conclusion of the study.

1.2 CHATBOT FOR CYBERSECURITY AND HOW IT WORKS

The word chatbot for cyber security has because an important aspect of studies, because it gives

us real-time support on security issues. Educating user on best way to prevent system and devices

from any form of attack, chatbot is good at responding and interaction to human because it was

design to mimic human. This study does not only enhance cyber security but also at as an educate

users on the precautions to take to avoid any form of attack

The goal of this research is to advance the field by creating a sophisticated Chabot that not only

acts as a proactive user educator but also acts as an advisor for information security. The objective

is to take advantage of chatbot beneficial effects in raising user awareness, combating the
ignorance of cyber-attacks, and consistently involving users in comprehending and mitigating

cyber threats.

This project attempts to develop a comprehensive solution that enables users to confidently

navigate the complicated world of cyber security and lack of user’s awareness in the field of cyber

security by expanding on the groundwork established by earlier studies and filling in the gaps

found in the body of existing literature.

1.3 CONCEPT OF ARTIFICIAL INTELLIGENT

The introduction of artificial intelligence has brought a massive advantage to the world

because artificial intelligence is a technology that mimic human behavior, AI focus on the

creation of system or algorithms that can perform tasks similar to human. Some task that

human perform such as solving problem, understanding language, making decision and

predictions. Artificial intelligence look forward to expand beyond human intelligence or

capabilities to perform some complex task. It perform some complex task that will take

human month to perform in minute (Wang, 2019). AI is generally classified into two

which is the weak and strong, Weak AI focuses on specific tasks, utilizing algorithms and

data to simulate intelligent behavior. For example, in the context of chatbot, Weak AI

enables these systems to mimic human conversation by responding to user inputs based

on pre-programmed rules and datasets, without any real understanding or consciousness

while Strong AI, by contrast, aspires to develop machines that can think, reason, and
understand like humans. While current AI systems, including chatbot, do not yet achieve

this level of sophistication, the long-term vision is to create AI that can engage in complex

and meaningful interactions, exhibiting true cognitive capabilities(Borana, 2016). It also

has a framework in which it operate on, those framework are the model used for the

development of a system which are the supervised large language model, unsupervised

learning model, reinforcement learning model and deep learning model and so on.

artificial intelligent can also be applied in various field of knowledge such as cyber

security, health field, chatbot, finance and educational aspect, it can also use for customer

services (Akhtar & Feng, 2021).

1.4 TECHNOLOGY OF CHATBOT

Using technologies like natural language processing (NLP), artificial intelligence (AI),

and machine learning (ML), chatbot also referred to as interactive agent or AI-powered

systems that mimic human conversation. ELIZA and other basic systems that employed

fundamental NLP techniques were early examples of chatbot evolution in the 1960s

(Hamad & Yeferny, 2020). According to (Lalpariya et al., 2022)chatbot of today are more

intelligent, apply powerful AI and ML algorithms to deliver tailored, dynamic responses.

Chatbot can be broadly classified into two types: AI-based chatbot, which use AI and ML

to understand and respond to a wide range of queries, learning from conversations to

improve over time (Hamad & Yeferny, 2020), and rule-based chatbots, which function on
predefined data and scripts, providing limited responses based on specific commands

(Lalpariya et al., 2022). Important technologies in chatbot creation include natural

language processing (NLP) and machine learning (ML), which allow chatbot to learn

from data and improve their responses over time (Hamad & Yeferny, 2020). NLP enables

chatbot to understand and process human language, stimulating meaningful

conversations (Lalpariya et al., 2022). Numerous advantages come with chatbot,

including improved customer service, continuously assistance, and lower costs for

operation. Nevertheless, there are certain limits that they must conquer. such as

managing complex queries and interpreting complex human emotions (Hamad &

Yeferny, 2020; Lalpariya et al., 2022).

2.0 RELATED WORK

Ranoliya et al. (2017) presented the use of chatbot in addressing the challenges of

enhancing of cyber security awareness and giving a solution that will support the users

having difficulties in security related issues. The study developed a chatbot using

OpenAI API. The chatbot was found to be efficient in providing real-time assistance and

increase learner’s knowledge of cyber security threats and also provide for guide for users

prevention from cyber threat.

According to Mondal et al. (2018), the issue of communication in the educational field

using chatbot to improve the communications. The system was developed using the
ensemble learning method and random forest algorithms to train the data and was

deployed as a telegram bot, assessable for users that need support on educational issues.

An emotional response generator by Lee et al. (2017) discussing system that understands

natural language and accurately interpret user emotional state using imputed data.

Multi-modal approach was employed to emotional recognition techniques with natural

language using probabilistic model to predict user emotion based on the user detail

inputted in the chatbot.

According to Gonda & Chu, (2019) a conversation bot to aid training courses was created,

highlighting the need for a training program that focuses on the gaps in graduate

teaching assistance to support users and also use technology to teach learning

experiences using collected data to evaluate the use of the graduate teaching assistance.

The use of chatbot has improved the support of graduate teaching assistants.

According to Van Heerden et al. (2017), the lack of access to HIV counseling and testing

services in countries with low income poses a concern in the health sector. The study

developed a chatbot that will guide users through an HIV counseling and testing session

during consultation.

The challenges of inefficient customer service spurred focused research into the

advancement of intelligent chatbot through SaaS solutions was addressed by D et

al.(2017). Which the study implemented the Amazon API, leveraging LUIS (Language
Understanding Intelligent Service) and Cognitive Services, to significantly enhance the

chatbot's natural language processing and communication efficacy.

The limited understanding of customer need and customer relationship, goggle was used

for survey of data, and the data was analyzed partial list square- structural equation

modelling via SMARTPLS 4.0. this improve service delivery Sofiyah et al.(2024).

Chatbot have proven to be effective tools for communication, simulating human

intelligence Roos, (2018). To address the limited application of chatbot in education,

developed an educational chatbot. By reviewing various studies and analyzing support

from several databases, the research focused on exploring how chatbot can enhance

education. This investigation identified key advantages and highlighted the need for

further research in this emerging field.

Colace et al., (2018) delves into the development of chatbot for e-learning has significantly

enhanced student support. By utilizing natural language processing and ontologies, these

chatbot generate automated responses, effectively supporting student learning. The

chatbot was evaluated and demonstrated strong performance in providing relevant

responses.

According to Lalpariya et al.(2022), a chatbot was developed to address issues with slow

college websites and difficulties in information retrieval by external users. This system

employed natural language processing and used a neural network created with the
tflearn library. A JSON file was prepared for training the model, improving its

effectiveness in resolving user queries.

During the COVID-19 pandemic, students encountered significant challenges in

communication and support. Balderas et al. (2023) addresses, a chatbot was developed

that utilizes language recognition and automated response technologies to provide

accurate and timely assistance to users.

Using a chatbot prototype, the SecBot project seeks to increase user involvement with

cyber security. The goal of the research is to enhance SecBot's security, usability, and

usefulness. In order to address chatbot vulnerabilities and help cyber security decision-

making, it highlights the necessity for continuous improvement, including improved

algorithms and tighter security safeguards Bulin Shaqiri, (2021).

The growing threat of hackers circumventing information system security is identified

by Michel et al. (2023), using a commercial SIEM system to train AI agents and install an

SMS alert system. Reports will be sent, security occurrences will be identified, and

general security will be monitored. But there are issues with the Elastic search engine's

weak correlation power and unexpected alarms.

The preservation and utilization of data in chatbot conversations are addressed by Hasal

et al. (2021), the data privacy in communication platforms and provide ways to safeguard
user data. The study draws attention to the cyber security dangers and ethical obligations

related to handling sensitive data in chatbot.

Al-Hawawreh et al. (2023) ChatGPT's application in cyber security, with a particular

emphasis on creating and executing False Data Injection (FDI) attacks against industrial

control systems (ICS) and supporting security analysts in thwarting cyber-attacks. They

draw attention to ChatGPT's capacity to generate FDI assault plans and scenarios.

Nonetheless, questions have been made concerning OpenAI's privacy policies,

specifically in relation to data transparency and the difficulties in using models trained

on generic data to handle domain-specific information.

Mohamed, (2023) focuses on the potential of AI and ML advancements in network

security, intrusion detection, and virus detection. The paper highlights the tremendous

potential of AI and ML for upcoming cyber security research and development while

identifying important obstacles, such as a shortage of qualified workers, a lack of

funding, and inadequate technology infrastructure.

Sebastian, (2023) identifies potential weaknesses that malevolent actors can exploit in

order to highlight the security hazards associated with AI chatbot such as ChatGPT.

Stronger safeguards and ongoing risk assessment are necessary in light of the constantly

changing tech ecosystem, even with current security measures. Although there is
enthusiasm for AI and ML among the public, there are also worries about data privacy

and social engineering attacks.

According to Arora et al. (2023), lowering vulnerabilities can improve chatbot security.

In order to combat cyber security risks, they implemented a chatbot on Twitter. While it

was successful in recognizing and reducing threats, it still has to be improved for greater

interaction and information extraction.

According to Sharma et al. (2023) IBM Watson Assistant was used to create "FLOKI," a

chatbot that teaches marine cadets collision avoidance procedures. After being tested on

the IBM Cloud and deployed on a Word Press website, the chatbot received positive

feedback and a SUS score of 73.72. The small sample size and regional variations in

attitudes toward technology are two of the study's acknowledged drawbacks.

According to Prasad & Ranjith, (2020)For lab security, the paper proposes a voice-based

chatbot that uses Automatic Speaker Recognition (ASR) for speech authentication. Like

virtual assistants like Alexa and Google Assistant, the device enhances lab security by

limiting access to authorize.

According (Gang et al., 2014)This study focus on the development chatbot of e-business

by using AIML or LSA in given a relevant and an immediate response to customer

questions. Unanswered questions are stored for future training. Evaluation achieved a
0.97 precision rate, demonstrating the effectiveness of combining AIML and LSA for e-

business customer service chatbots.

Sousa et al. (2019) focused on enhancing customer support by developing a taxonomy

that compares various chatbot technologies, such as Microsoft Bot Framework and

DialogFlow. This study aimed to assist enterprises in selecting the most suitable solutions

by highlighting the natural language integration capabilities of each technology. The

research revealed that custom-built solutions often surpass commercial products in

performance. The findings suggest that employing a structured approach can help

organizations effectively navigate the complexities of chatbot technology, leading to

improved user interaction and satisfaction.

Li and Ye, (2020) Vulnerabilities in client, communication, response generation, and

database modules are the main areas of attention for this research concerning chatbot

architecture security. It recognizes that industry standards are lacking and that ongoing

security evaluations are necessary, and it emphasizes the necessity of constant study due

to increasing threats.

"Chabot Sec" (CBS), a chatbot that offers guidance on information security, is introduced

in the paper by Hamad & Yeferny, (2020) CBS enhances the dependability of information

supplied to consumers by employing keyword extraction to find pertinent terms during

user interactions and retrieving precise guidance from a knowledge base based on JSON.
Gupta et al. (2023) the study focuses on how Gen AI technologies are used in cyber

security, emphasizing both offensive and defensive strategies. In addition to exploring

difficulties with automating cyber security, threat intelligence, and safe code production,

it tackles social, ethical, and privacy issues. The report recommends more investigation

to maximize the use of Gen AI technologies in cyber security.

The study of Wang et al., (2023) discuss the ethical, legal, security, and privacy concerns

related to AI-generated content (AIGC), including ChatGPT. The essay emphasizes how

important it is to put in place safe frameworks and laws to mitigate potential risks. It also

demonstrates the limitations of current research and the rapid development of AIGC

technology.

Wu et al. (2023) analyzes ChatGPT, focusing on its development, characteristics, and

ethical considerations. It highlights limitations such as inaccuracies, security risks,

privacy breaches, and biases, along with challenges in detecting AI-generated content for

plagiarism. The study emphasizes the need for robust detection mechanisms to address

these issues.

The study Addington, (2023), focus on hazards of information leakage, phishing, and

manipulation related to ChatGPT are covered in this study. To lessen these hazards,

OpenAI has put safety measures in place including data encryption and security

monitoring.
Abdullahi et al.,(2022), delves into the use of deep learning (DL) and machine learning

(ML) methodologies for IoT security, with a specific focus on identifying and addressing

cyber security risks. It emphasizes how crucial AI-powered sophisticated intrusion

detection systems (IDS) are. Many IoT frameworks still lack strong procedures, which

causes vulnerabilities and erroneous results despite developments.

Das and Sandhane (2021) focuses on an enormous improvement in threat management

was made possible by the application of AI techniques such as neural networks to address

the complexity of cyber operations and strengthen security defenses. Expert systems and

data mining are two other AI techniques that have been used to address the absence of

intelligent systems in sectors like gaming and weather forecasting. All these areas saw

increased decision-making and efficiency as a consequence. Due to its ability to automate

intricate procedures and enhance results, artificial intelligence has the potential to

revolutionize a wide range of sectors. AI has shown itself to be an invaluable instrument

for a variety of technological applications, including cyber security.

Borana, (2016) Artificial intelligence techniques like neural networks have been

successfully used to address the complexity of cyber operations and strengthen security

defenses, resulting in notable advancements in threat detection and management.

Furthermore, the use of AI techniques like data mining and expert systems in sectors like

gaming and weather forecasting has enhanced productivity and improved the quality of

decision-making. In general, the incorporation of artificial intelligence technologies into

diverse sectors has been crucial in resolving intricate issues and enhancing efficiency,

showcasing the extensive possibilities and adaptability of AI in contemporary uses.

Mohammad Aniq Bin Amdan et al.(2024) by utilizing AI abstractions like goals and

commitments, the rigidity of chatbot constructed using If-This-Then-That frameworks

was addressed, leading to increased flexibility and maintainability. Furthermore, goal-

based dialog management and AI-driven chatbot were used to address the complexity of

IT service administration, resulting in increased automation and better service quality.

de Bruijn & Janssen, (2017) emphasizes the urgent need for increased public awareness

of cyber security, which is still lacking despite its crucial importance. It looks at the

complexities of the cyber security discourse, which often falls short of engaging a range

of audiences and leads to an inadequately proactive response to possible threats. The

authors support evidence-based framing techniques that make cyber security problems

more approachable and urgent. This will guarantee that different groups of people

individuals as well as businesses get tailored messages that speak to their specific

concerns and experiences. The goal of this approach is to increase political and societal

awareness of the complex issues that cyber security presents.

(Li et al., 2019) reveals that as cyber threats increase, employee compliance with

information security policies is crucial; it found that awareness of these policies

significantly enhances employees' ability to manage cyber security tasks, with results
showing that a supportive organizational environment positively influences employees'

threat appraisal and compliance behavior, ultimately reducing the risk of security

breaches in organizations.

Blease, (2023) discuss the challenges patients face in understanding their electronic health

records due to complex medical terminology and the traditional focus of clinical records

on clinician communication rather than patient accessibility. It examines the implications

of open record access (ORA) and surveys both patient and clinician experiences,

highlighting that while patients often feel empowered and more informed by accessing

their records, significant privacy concerns arise regarding the exposure of sensitive

information to technology companies. These insights underscore the need for improved

communication and privacy safeguards in the context of ORA.

The integration of AI in education faces challenges like academic integrity issues, which

were addressed through the analysis of various studies revealing both positive effects

and concerns, while personalization of education was enhanced by machine learning,

resulting in improved student engagement and efficiency in administrative tasks through

AI automation Chen et al., (2020).

As stated by Chandarman & Niekerk, (2017) Cyber security awareness has been an area

of focus due to the increase in cyber-attack and lack of cyber awareness, the study aimed

to assess the levels of cyber security awareness among students of a private tertiary
education institution in South Africa. Data were collected via a questionnaire and was

analyzed using the SPSS statistical software package.

Erendor & Yildirim, (2022) assessed cyber security awareness among students at Kyrgyz-

Turkish Manas University during online education. A survey revealed low awareness of

cyber security threats, such as phishing and malware, with male and graduate students

showing higher knowledge levels. The results highlight the need for cyber security

education to enhance students' ability to protect against cyber-attacks, particularly in an

increasingly digital learning environment

According to Quayyum et al. (2021) the increasing cyber security risks for children due

to their internet exposure. The methodology involved a systematic literature review

analyzing studies on cyber security awareness for children, focusing on risks and

awareness-raising approaches. The results highlighted various cyber security risks,

approaches for raising awareness, and factors influencing the evaluation of these

approaches,

The study addresses barriers to cyber security information sharing, focusing on legal,

performance, technological, and managerial challenges that deter firms from sharing

data. A survey was conducted among cyber security professionals to identify these

barriers and incentives for sharing. Results indicated that legal concerns were the most

significant deterrent, with 56% of respondents highlighting this issue(Koepke, 2017)

3.0 Methodology
The method adopted in this project involves building a Chabot that would raise

awareness about cyber security. The first phase entails a thorough analysis of the body of

research on cyber security awareness and Chabot functionality. Its goals and scope are

therefore clearly stated, and the project places a strong emphasis on building a Chabot

that instructs users on basic cyber security procedures.

3.2 DESIGN
The system architecture for this project shown in figure 3.1involves multiple components

working together to provide a seamless user experience. The user experiences a request for input

into the chatbot which is expected to be a cyber security related question, after inputting the

question the chatbot will check if the question is cyber security related, if yes then it will process

it using (OpenAI: GPT 3.5) and send the output then asked if the user will like to ask another

question. While, if the question is not cyber security related questions, the system will output a

query “Please enter a cyber security question only”.

Figure 1: The architectural design of the project

3.3 SYSTEM COMPONENT
The system component of the user interface consists of the language mode which is the (OpenAI:

GPT 3.5), the keyword knowledge-based and natural language. Each of the components plays a

critical role in ensuring the chatbot provides personalized real-time assistance, adapts to user

understanding, and maintains data accuracy through the learning processes.

Language model: the utilization of the OpenAI’s GPT 3.5 as the core language model for natural

language understanding and generation

Natural language processing (NLP): The implementation of natural language processing for

intent recognition and entity extraction

Knowledge base: a structured repository of cyber security information, including threat

intelligence, and best practices was created.

To respond to threat intelligence, phishing detection, malware analysis, and so on, this project

employs an OpenAI model. To receive an accurate and up-to-date response, a new request is sent

to the API for fresh questions asked by the user.

3.1. CRAWLING AND DATA ACQUISITION

Data crawling is the first stage in creating the AI-enhanced dark web surveillance system.

This entails harvesting data from the dark web using web scraping techniques. Only
specialist software, such as TOR, can grant access to the dark web, a hidden portion of

the internet. A Node.js software called Puppeteer provides a high-level API for managing

headless Chrome or Chromium browsers. It was employed in this project as seen in

Figure 6, the headless browser will launch using the default SOCKS proxy address for

TOR in order to successfully crawl the dark web. Without this, it will be impossible to

successfully crawl the dark web using our default browser proxy address because access

to the dark web requires specialized software like TOR. This will crawl the whole body

of the onion site and the extracted data is stored in JSON format for further processing.

Each crawled onion site content is saved as a JSON object, enabling structured data

storage and easy retrieval for subsequent stages.

3.4 CHATBOT USER INTERFACE:

HTML is the foundation for organizing the structure and elements of this project. Text,

graphics, forms, and other interactive components were among the elements of the user

interface (UI) that were defined using HTML. These components provide data entry,

application navigation, and feature interaction for users. A clean, well-structured

framework made possible by the incorporation of HTML guarantees fluid user

interaction and content presentation. CSS and JavaScript are also utilized in this project.

CSS is used to style and layout the Chabot user interface, making it visually appealing

and ensuring a consistent design across different devices and screen sizes. JavaScript is

used to add interactivity and functionality, allowing the Chabot to process user inputs,
respond in real time, and integrate with other services. It is also used to interact with the

APIs of the large language mode (OpenAI: GPT-3.5). The request from the chatbot to the

API is in JSON format. The API also responded in JSON format.

4.0 RESULT AND DISCUSSION:

This chapter contains the results from the creation and assessment of intelligent cyber security.

The chatbot has shown noteworthy advancements in augmenting cyber consciousness,

identifying potential hazards, and imparting cyber security knowledge to its users. The key to

obtaining high accuracy and successful threat identification was the API-based integration of the

OpenAI model. The favorable results and user comments show that the Chatbot is an effective

instrument for strengthening cyber security defenses and swiftly addressing threats.

4.1 INTERFACE DEVELOPMENT

To make using the API easier, an interface was created. The user interface enables users to the

interface facilitates communication between the user and the Chabot that transmits questions to

the "API" for response generation. The user can type questions into the interface and the Chabot

will forward them to the "API," which will process the data through a model and display the

results on the website. The system's usability and accessibility for cyber security awareness are

improved by its intuitive interface.

4.2 USER EDUCATION AND THREAT DETECTION

The chatbot provided users access to up-to-date knowledge and best practices for cyber security;

after interacting with it, users' comprehension of cyber security-related topics increased; Users
may ask questions and receive thorough explanations on a variety of cyber security subjects

during interactive learning sessions that were supported by the system.

An impressive of threats were detected accurately as a result of the OpenAI model's integration,

which also made sure the system always had the most recent threat intelligence, enhancing

detection skills overall.

4.3 PERFORMANCE EVALUATION

This is the performance review of the chatbot to make sure it is effective and operates

flawlessly. The timeliness and precision of the message's response

4.3.1 The accuracy of the chatbot: The chatbot's accuracy in intent detection shows that it can

recognize user intents in the majority of exchanges. This was assessed by contrasting the chatbot's

answers with a list of pre-established intents that were derived from user inquiries.

4.3.2 Response time: Less than 5 seconds is the average reaction time for the chatbot, which is

deemed effective for in-the-moment engagements. The user experience and engagement are

improved by this prompt reaction.

4.4User Interaction with the Chatbot

The interactions between users and the intelligent chatbot created for cyber security are shown in

the following images. These exchanges demonstrate the chatbot's ability to deliver pertinent

information promptly while maintaining a user-friendly interface.

Figure 2 A conversion between the user and the chatbot on how to prevent hacking
This is one of the examples of an interaction between the chatbot and the user in the figure above

in which a question was inputted by the user asking how to prevent hacking and the chatbot

responded and gave an accurate response on how to prevent system or devices from getting

hacked by a hacker.

Figure 3 A conversation with chatbot educating the user on the hacking

The chatbot was developed to also educate users on cybersecurity questions, therefore the chatbot

serves as an education that educates users on cyber-related issues. Giving detailed answers to

every question asked by the user. Shown figure 3 how a user asks the chatbot what is and the

chatbot gives the user an accurate response on hacking.

4.5 DISCUSSION
The findings show how intelligent the chatbot can raise user awareness of cyber security issues

and offer them real-time support. Positive feedback was given by the user and the chatbot's strong

intent recognition and entity extraction accuracy demonstrate how successful it is as a tool for

spreading cyber security awareness. To educate users on threats and how users can overcome

them, however, continual refinement based on user interactions and feedback will be done.

4.6 CONTINUOUS LEARNING AND ADAPTATION

Based on the API that was used to create the chatbot, the knowledge base of the chatbot will be

updated regularly with the most recent cybersecurity risks and best practices. This guarantees
that the chatbot will always be up-to-date and capable of giving precise answers to any

cybersecurity-related queries.

4.7 LIMITATIONS
The chatbot, which utilizes the OpenAI API, excels at answering cyber security-related questions

but has trouble answering more complicated or ambiguous ones. Performance problems

occasionally resulted from integration with the API.

5.0 SUMMARY AND CONCLUSION

The incredible potential of artificial intelligence (AI) in the field of cyber security was

brought to light by the development and implementation of the chatbot using OpenAI's

API. The chatbot's efficiency in responding to cyber security-related inquiries illustrated

the benefits of utilizing AI to address intricate and dynamic security challenges. It also

proved to be a useful tool for raising cyber security awareness, offering precise and timely

assistance, and assisting users in understanding and mitigating various cyber threats.

However, the project also revealed several limitations.

The chatbot needed more extensive training and advanced natural language processing

skills because its performance in responding to questions unrelated to cyber security was

less reliable. Furthermore, user feedback indicated that the user interface needed to be

improved even further, and that the chatbot needed to be able to handle more

sophisticated queries and a larger range of themes.

The experiment was successful overall and demonstrated how AI can revolutionize cyber

security. The results indicate that chatbots driven by artificial intelligence (AI) have the

potential to significantly improve cyber security education and offer invaluable

assistance, even though there is still space for growth. Future development should

concentrate on enhancing the chatbot's natural language comprehension, increasing its

skills, and making sure it can continue to adapt to the always-shifting cyber security

threat scenario.

5.3 RECOMMENDATION

In the future, the chatbot needs to be updated because it uses OpenAI's API to gain

responses. The OpenAI model gets updated time-to-time which means the OpenAI

advances and adds more features to its model. To make it more efficient in answering

more advanced cyber-related issues. So, therefore, the need for updating the model used

for the Chatbot from the OpenAI will be necessary for the chatbot to meet up with the

new threat and how to go about it. Lastly Enhancing the Chatbot's user interface will

make it easier to use. Expanded field testing and user input integration will aid in finding

and resolving problems.

REFRENCES

Abdullahi, M., Baashar, Y., Alhussian, H., Alwadain, A., Aziz, N., Capretz, L. F., & Abdulkadir,
S. J. (2022). Detecting Cybersecurity Attacks in Internet of Things Using Artificial
 Intelligence Methods: A Systematic Literature Review. Electronics (Switzerland), 11(2), 1–
27. https://doi.org/10.3390/electronics11020198

Addington, S. (2023). ChatGPT: Cyber Security Threats and Countermeasures. SSRN Electronic
Journal, 1–12. https://doi.org/10.2139/ssrn.4425678

Akhtar, M., & Feng, T. (2021). An overview of the applications of Artificial Intelligence in
Cybersecurity. EAI Endorsed Transactions on Creative Technologies, 8(29), 172218.
https://doi.org/10.4108/eai.23-11-2021.172218

Al-Hawawreh, M., Aljuhani, A., & Jararweh, Y. (2023). Chatgpt for cybersecurity: practical
applications, challenges, and future directions. Cluster Computing, August.
https://doi.org/10.1007/s10586-023-04124-5

Bay, M. (2016). What is Cybersecurity? In search of an encompassing definition for the post-
Snowden era. French Journal For Media Research, Special Is(6), 1–28.

Blease, C. (2023). Open AI meets open notes: surveillance capitalism, patient privacy and online
record access. Journal of Medical Ethics, 50(2), 84–89. https://doi.org/10.1136/jme-2023-
109574

Borana, J. J. N. U. (2016). Applications of Artificial Intelligence & Associated Technologies.

Proceeding of International Conference on Emerging Technologies in Engineering,
Biomedical, Management and Science [ETEBMS-2016], March, 25–30.

Bulin Shaqiri. (2021). Development and Refinement of a Chatbot for Cybersecurity Support.
Https://Www.Merlin.Uzh.Ch/Publication/Show/20879.
https://www.merlin.uzh.ch/publication/show/20879

Chandarman, R., & Niekerk, B. Van. (2017). Students ’ Cybersecurity Awareness at a Private
Tertiary Educational. The African Journal of Information and Communication (AJIC), 20,
133–155.

Chen, L., Chen, P., & Lin, Z. (2020). Artificial Intelligence in Education: A Review. IEEE Access,
8, 75264–75278. https://doi.org/10.1109/ACCESS.2020.2988510
Colace, F., De Santo, M., Lombardi, M., Pascale, F., Pietrosanto, A., & Lemma, S. (2018). Chatbot
for e-learning: A case of study. International Journal of Mechanical Engineering and
Robotics Research, 7(5), 528–533. https://doi.org/10.18178/ijmerr.7.5.528-533

Craigen, D., Diakun-Thibault, N., & Purse, R. (2014). Defining Cybersecurity. Technology
Innovation Management Review, 4(10), 13–21. https://doi.org/10.22215/timreview835

Dahiya, M. (2017). A Tool of Conversation : Chatbot International Journal of Computer Sciences

and Engineering Open Access A Tool of Conversation : Chatbot. December.

Das, R., & Sandhane, R. (2021). Artificial Intelligence in Cyber Security. Journal of Physics:
Conference Series, 1964(4). https://doi.org/10.1088/1742-6596/1964/4/042072

de Bruijn, H., & Janssen, M. (2017). Building Cybersecurity Awareness: The need for evidence-
based framing strategies. Government Information Quarterly, 34(1), 1–7.
https://doi.org/10.1016/j.giq.2017.02.007

Erendor, M. E., & Yildirim, M. (2022). Cybersecurity Awareness in Online Education: A Case
Study Analysis. IEEE Access, 10, 52319–52335.
https://doi.org/10.1109/ACCESS.2022.3171829

Gang, W. Y., Bo, S., Chen, S. M., Yi, Z. C., & Zi, M. P. (2014). Chinese intelligent chat robot
based on the AIML language. Proceedings - 2014 6th International Conference on Intelligent
Human-Machine Systems and Cybernetics, IHMSC 2014, 1, 367–370.
https://doi.org/10.1109/IHMSC.2014.96

Ghorashi, N., Ismail, A., Ghosh, P., Sidawy, A., & Javan, R. (2023). AI-Powered Chatbots in
Medical Education : Potential Applications and Implications. 15(8), 1–5.
https://doi.org/10.7759/cureus.43271

Gonda, D. E., & Chu, B. (2019). Chatbot as a learning resource? Creating conversational bots as
a supplement for teaching assistant training course. TALE 2019 - 2019 IEEE International
Conference on Engineering, Technology and Education.
https://doi.org/10.1109/TALE48000.2019.9225974
Hamad, S., & Yeferny, T. (2020). A chatbot for information securityHamad, S., & Yeferny, T.
(2020). A chatbot for information security. ArXiv, 20(4), 287–291. ArXiv, 20(4), 287–291.

Hiremath, G., Aishwarya, H., Priyanka, B., & Nanaware, R. (2020). Chatbot for education system.
International Journal of Advance Research, Ideas and Innovations in Technology, 4(3), 37–
43.

Jafarov, S. (2023). AI AND CYBERSECURITY -NEW THREATS AND OPPORTUNITIES.

December.

Koepke, P. (2017). Cybersecurity Information Sharing Incentives and Barriers-Massachusetts

Institute of Technology. Cybersecurity Interdisciplinary Systems Laboratory (CISL), 2017–
13(June), 1–38. http://web.mit.edu/smadnick/www/wp/2017-13.pdf

Lalpariya, M. M., Murad, K., Adil, M., Amin, S. M., & Akbar, K. A. (2022). Ai Chatbot. 4(6),
1107–1114. https://doi.org/10.35629/5252-040611071114

Lee, D., Oh, K. J., & Choi, H. J. (2017). The chatbot feels you - A counseling service using
emotional response generation. 2017 IEEE International Conference on Big Data and Smart
Computing, BigComp 2017, 437–440. https://doi.org/10.1109/BIGCOMP.2017.7881752

Li, L., He, W., Xu, L., Ash, I., Anwar, M., & Yuan, X. (2019). Investigating the impact of
cybersecurity policy awareness on employees’ cybersecurity behavior. International Journal
of Information Management, 45(February 2018), 13–24.
https://doi.org/10.1016/j.ijinfomgt.2018.10.017

Michel, D. D. E., Emmanuel, T., Franck, B. S. F. A., & Alphonse, B. A. (2023). Design and
Implementation of a Chatbot for the Supervision of Security Events (SIEM). International
Journal of Computer Applications, 185(14), 41–53. https://doi.org/10.5120/ijca2023922831

Mohamed, N. (2023). Current trends in AI and ML for cybersecurity: A state-of-the-art survey.

Cogent Engineering, 10(2). https://doi.org/10.1080/23311916.2023.2272358

Mohammad Aniq Bin Amdan, Naldo Janius, & Mohd Aidil Hazidi Bin Kasdiah. (2024). Concept
paper: Efficiency of Artificial Intelligence (AI) tools For STEM Education In Malaysia.
 International Journal of Science and Research Archive, 12(2), 553–559.
https://doi.org/10.30574/ijsra.2024.12.2.1273

Mondal, A., Dey, M., Das, D., Nagpal, S., & Garda, K. (2018). Chatbot: An automated
conversation system for the educational domain. 2018 International Joint Symposium on
Artificial Intelligence and Natural Language Processing, ISAI-NLP 2018 - Proceedings, 1,
1–5. https://doi.org/10.1109/iSAI-NLP.2018.8692927

Prasad, V. A., & Ranjith, R. (2020). Intelligent Chatbot for Lab Security and Automation. 2020
11th International Conference on Computing, Communication and Networking Technologies,
ICCCNT 2020, 1–4. https://doi.org/10.1109/ICCCNT49239.2020.9225641

Quayyum, F., Cruzes, D. S., & Jaccheri, L. (2021). International Journal of Child-Computer
Interaction Cybersecurity awareness for children : A systematic literature review.
International Journal of Child-Computer Interaction, 30, 100343.
https://doi.org/10.1016/j.ijcci.2021.100343

Ranoliya, B. R., Raghuwanshi, N., & Singh, S. (2017). Chatbot for university related FAQs. 2017
International Conference on Advances in Computing, Communications and Informatics,
ICACCI 2017, 2017-Janua, 1525–1530. https://doi.org/10.1109/ICACCI.2017.8126057

Shamiulla, A. M. (2019). Role of Artificial Intelligence in Cyber Security. 1, 4628–4630.

https://doi.org/10.35940/ijitee.A6115.119119

Sofiyah, F. R., Dilham, A., Hutagalung, A. Q., Yulinda, Y., Lubis, S., & Marpaung, J. L. (2024).
The chatbot artificial intelligence as the alternative customer services strategic to improve
the customer relationship management in real-time responses The chatbot artificial
intelligence as the alternative customer services strategic to improve the cus. January.
https://doi.org/10.1504/IJEBR.2024.10064925

Van Heerden, A., Ntinga, X., & Vilakazi, K. (2017). The potential of conversational agents to
provide a rapid HIV counseling and testing services. Conference Proceedings - 2017
International Conference on the Frontiers and Advances in Data Science, FADS 2017, 2018-
Janua, 80–84. https://doi.org/10.1109/FADS.2017.8253198
Wang, P. (2019). On Defining Artificial Intelligence. Journal of Artificial General Intelligence,
10(2), 1–37. https://doi.org/10.2478/jagi-2019-0002