Integration of Natural Language Processing Chabot’s into

Web Platforms for Intelligent Customer Support

Mrs. Rupinder Kaur, Dr. Hemant Kasturiwale, Mr. Varunkumar Mishra, Mr. Sumit Kumar
Thakur College of Engineering and Technology Mumbai, India
rupinder.kaur@tcetmumbai.in,Hemant.kasturiwale@tcetmumbai.in,varun.mishra@tcetmumbai.in,sumit.kumar@tcetmumbai.in

Abstract—The design and implementation of a decentralized AI-powered machines to derive meaningful information from human
chatbot that is incorporated into a web platform using advanced Natural language inputs by analyzing textual data.
Language Processing (NLP) techniques to provide safe, contextually
aware, and customized user interactions is presented in this paper. With its It processes language based on its hierarchical structure, where
decentralized architecture, encryption, and secure API connectivity, the words form phrases, phrases construct sentences, and sentences
system—which is based on the Gaia Net blockchain infrastructure— convey meaningful ideas [2]. While enabling computers to
ensures data privacy, integrity, and scalability. In order to properly understand human language offers significant advantages, AI
understand user intent and produce pertinent, human-like responses— researchers often underestimate the inherent complexity involved
while integrating user history and preferences for improved engagement— in both the generation and comprehension of natural language.
the chatbot makes use of Natural Language Understanding (NLU) and The challenge extends beyond understanding individual words to
Natural Language Generation (NLG). Requirement analysis, NLP model
grasp their contextual usage, expressions.[3]
generation, RESTful API connection, and thorough testing were all
covered by an organized development process. excellent accuracy Organizations across various sectors increasingly adopt Chabot
(92.3%), precision (90.8%), recall (93.6%), and quick response times (600 systems, often in the form of virtual agents, to assist users via
ms) were demonstrated in the performance test, along with excellent user desktop and web-based platforms. An ideal Chabot’s should
satisfaction across all demographic categories. While highlighting possess the ability to learn from interactions, continuously
potential future improvements like multilingual support, emotional improve its performance, and maintain contextual understanding,
sensitivity, and on-device learning for better interaction quality, the ensuring more effective and meaningful user engagements. [4]
suggested system shows how AI, NLP, and blockchain technologies can
be combined to create intelligent, secure, and adaptive conversational
agents for a variety of applications. II. LITERATURE SURVEY
Keywords— Natural Language Processing, Chatbot, Artificial Chatbots, which are frequently denoted by alternative terms such
Intelligence, Secure data handling. as talk-bots or conversational agents, represent sophisticated
computer-based systems meticulously engineered to replicate
I.INTRODUCTION human-like interactions, predominantly through various online
Among this AI powered chatbots offering instant and round-the- communication platforms that are integral to modern digital
clock support across various digital platforms. Their integration into communication. The inception of the concept of chatbots can be
web-based systems not only simplifies communication processes traced back to the early 1990s, specifically when Michael Mauldin
but also enhances user engagement by delivering personalized and pioneered the development of "Chatterbot," which stands as a
efficient interactions. NLP plays a vital role in empowering these notable milestone as the first practical implementation of such
chatbots by enabling them to comprehend, interpret, and respond to systems, a significant achievement that occurred in 1994[5].
human language naturally
This Turing Test involved three distinct participants — a man, a
and intuitively. This capability allows chatbots to address a woman, and an interrogator — and was designed with the specific
diverse range of user queries, from routine frequently asked objective of determining whether the interrogator could
questions to complex customer service requests, making them successfully differentiate between the two human participants
valuable assets across industries such as e-commerce, healthcare, based solely on their responses to questions posed, without any
education, and beyond.With the growing demand for instant additional contextual cues [7,6].
assistance and tailored services, businesses increasingly recognize
the necessity of incorporating AI chatbots powered by NLP into Entities across diverse industries are progressively integrating
their platforms to enhance customer satisfaction and optimize chatbot systems, frequently manifested as virtual agents, to
operational workflows. Such integration helps automate repetitive facilitate user assistance through desktop and web-based
tasks, reduces response times, and allows human agents to focus on interfaces. The ascendance of chatbots is attributable to their
more complex queries, ultimately contributing to improved user perpetual availability, around-the-clock support capabilities,
experiences and organizational productivity [1]. and the convenience they provide in promoting seamless human
computer interaction. As a result, it is projected that chatbots
The continuous advancements in AI bring humanity closer to a
may ultimately supplant traditional conversational interfaces.
future where machines can address complex challenges
Moreover, with the escalating proliferation of mobile
independently A chatbot is typically defined as a software
devices, chatbots are increasingly prevalent owing to their
application that leverages NLP techniques to interact with users
cognitive intelligence, adaptability, and user-centric design.
and address their needs. NLP, as a branch of AI, enables An exemplary chatbot should exhibit the capacity to learn
from its interactions, consistently enhance its performance,
and sustain contextual awareness, thereby ensuring more configured correctly. Monitoring tools are implemented to track
impactful and meaningful user engagements. performance and user interactions, along with a maintenance
A decade subsequent to the initial development, the advent of plan for ongoing updates based on user feedback. Finally, an
ELIZA represented a pivotal advancement within the domain of evaluation and iteration phase allow for performance analysis
artificial intelligence. ELIZA emerged as the inaugural publicly and continuous user feedback, ensuring that the chatbot evolves
accessible chatbot, proficient in conducting dialogues through to meet user needs effectively. This structured methodology aims
pre-established scripts. Despite the constraints of its to create a robust, effective, and user-friendly Chabot that
conversational proficiency, ELIZA successfully emulated enhances interactions and delivers valuable services through
substantive interactions by channeling user inputs via generalized advanced NLP technologies.
frameworks; AI scholars frequently underestimate the intrinsic IV. DESIGN
intricacy associated with both the production and interpretation of Because it dives more profound into the subject, the Chabot starts by
natural language [8]. The quandary transcends mere looking at the essential themes. It employments the pipe concept to
comprehension of isolated lexemes to encompass the take after social, behavioral, and mental angles in arrange to center
understanding of their contextual application, idiomatic on the essential subject. In conclusion, a rundown of the
expressions, and collocational patterns. communication comes about between the client and Chabot is given.
Organizations across various sectors increasingly adopt chatbot Individuals believe Chabot more when they see it as a group player
systems, often in the form of virtual agents, to assist users via instead of as a mechanical instrument [18,19].
desktop and web-based platforms. The popularity of chatbots Data is seen as honest to goodness when chatbots talk so also to how
continues to grow due to their continuous availability, 24/7 their clients do. The writing on brain research and human science
support capabilities, and the convenience they offer in facilitating addresses orientation generalizations.
seamless human-computer interaction. Consequently, it is For occurrence, buyers are more certain in a male Chabot than in a
anticipated that chatbots may eventually replace conventional female one when it comes to specialized questions. Individuals
conversational interfaces in many applications [9]. Furthermore, ordinarily just like the female chatbots when asking help from client
with the increasing proliferation of mobile devices, chatbots are benefit centers. Chatbots lose believe in case they inquire the same
becoming more prevalent due to their intelligence, adaptability, questions over and over once more and give clients destitute or
and user-friendly nature. An ideal chatbot should possess the erroneous reactions [20].
ability to learn from interactions, continuously improve its This makes clients not utilize them any longer. Bots ought to express
performance, and maintain contextual understanding, ensuring positive feelings like delicacy, bliss and delight to construct up the
more effective and meaningful user engagements [10]. relationship and engagement with the clients. A great chatbot ought
to express tiniest level of compassion and ought to carry on caring
In the later years, ALICE building upon the foundational and situational. Chatbot can too lock in in little talk and amusement.
concepts established by ELIZA. Unlike its predecessors, ALICE Chabot is favored in the event that it doesn't continuously react
was released as open- source software, which allowed quickly but coordinating certain stops a bit like human creatures do
developers to extend its dialogue patterns and improve its [21].
conversational capabilities [11,12].
III. TECHNOLOGY
The methodology for developing the AI Chatbot Integration
project using NLP is structured to ensure a systematic approach
through several key phases.
The project begins with comprehensive planning, which includes
gathering requirements from stakeholders through surveys and
interviews, conducting a feasibility study to evaluate the project's
viability, and clearly defining the scope and objectives. Next, the
design phase focuses on creating the architecture, user interface,
and the NLP model, incorporating feedback from potential users
to refine the design [13,14]. During the implementation phase,
the development environment is set up with the chosen
technology stack, followed by the creation of the NLP model and
the development of RESTful APIs for seamless interaction
between the chatbot and the web application. Testing follows,
encompassing unit tests for individual components, integration
tests to ensure components work together, and user acceptance
testing (UAT) to validate the chatbot's performance in real-world
scenarios [15,16]. Once testing is complete, the deployment Fig 1: Decentralized Design Framework
phase involves executing a well-planned strategy to launch the
chatbot on a cloud platform while ensuring all components are
 personalized interactions. At the Presentation Layer, users
interact via a web or mobile interface, admin dashboards, and
optional node status panels. The Application Layer has center
functionalities like lightweight Chabot rationale Phi-3 Scaled
down, domain-specific information, discussion history,
specialist personalization, and session administration. The
Data Layer supports model checkpoints, OpenAI-compatible
APIs, and optional off-chain storage. The Integration Layer
connects the Chabot with the Gaia net node API, external AI
APIs, and optionally IPFS for static asset storage. At long
final, the Security Layer ensures end-to-end confirmation
through wallet or token based declaration, on-device or node-
Fig 2: Use Case Diagram level encryption, and on/ off-chain think about logging to
secure data insightfulness and client recognize.

V. METHODOLOGY
The present research emphasizes the development of chatbots
utilizing Natural Language Processing (NLP) techniques,
highlighting the indispensable role of NLP in enabling
chatbots to interpret human speech and effectively convert
Fig 3: Top Level Data Flow Diagram spoken input into textual form. The foundation of this process
The chatbot design and interaction is based on the formulation lies in two critical components of NLP: Natural Language
Understanding (NLU) and Natural Language Generation
of R=f(M(x), G(n), P(u)) (NLG) [22]. These let chatbots understand human language
The chatbot's final response R is a result of the AI model's output and construct appropriate meaning in response to given text.
M(x), processed through Gaia Net’s decentralized execution Chatbots based on NLP are built using multiple methods,
such as deep learning algorithms, structure-based pattern
environment G(n), and optionally influenced by user-specific context
recognition, AIML (Artificial Intelligence Markup
P(u).
Language), and syntactic parsing [23]. Pattern matching
contributes greatly to the development of automated agents
that fetch information for users, provide customer service,
solo dependancy systems, or virtual assistants, like the
popular Siri or Apple virtual assistant, and other similar
intelligent virtual agents.
Additionally, this research expands on some of the most basic
NLP methods useful in crafting a chatbot. One of the most
important steps is Part-of-Speech (POS) tagging, where
individual tokens or words are divided into several
grammatical categories. These include the nouns, verbs,
adjectives and others which assist interpretation [24].
Stemming also does the same by reducing Inflected or derived
words to a base form for specific rules and lemmatization
considers context to determine the correct lemma. Other
important techniques such as entity recognition and do named
entity classification aid in extracting information from user
data and understanding the constituent parts of a sentence.
Together, these basic assistant techniques improve the ability
of the chatbot to ensure proper contextual understanding and
Fig 4: Level 1 Data Flow Diagram response generation.[25]
VI. BLOCK DIAGRAM & FLOWCHART
This encapsulates both the intelligence of the chatbot and the A chatbot’s incorporation into a web application is
decentralized routing, privacy, and compute handled by Gaia hierarchical, methodical, and guarantees optimal user
Net. where x is the user input.
experience. This process starts with the user entering details
As seen in figure 1, This layered architecture represents a regarding the chatbot using the GUI provided in the web
decentralized chatbot system built for secure, scalable, and application. The user's input is sent to a web server, which
 forwards the message to the chatbot engine. The NLP powered remarkable accuracy and relevance in a timely manner. As
chatbot engine processes the user query, identifies the intent, users noted, the chatbot's answers were contextually aware,
and formulates a response accordingly [26]. If more informative, and responded to the user which made interactions
information is required, the chatbot can access backend more captivating and satisfying. The backend architecture and
databases or other APIs to supplement existing data. It will NLP integration of the chatbot proved effective as the smooth
thus access these databases or APIs to generate relevant interactions throughout busy periods validate peak
response[27].By reaching into the world of the internet, with performance. The responsive design provided the surety that
their words the user can engage the chatbot system and ask a the application would be scalable and could adjust to different
question. Backend server receives this query and connects the windows as well as devices which could be applied in
frontend to the chatbot engine. How does this work? Through Matterhorn.
the use of an embedded NLP processor, the chatbot decodes The following results were recorded from different components
input data, uncovers relevant information, and identifies the •Accuracy: 92.3%
user's objective. The chatbot may use pre-defined •Precision: 90.8%
conversational rules, machine learning models, or external •Recall: 93.6%
API calls to generate a suitable answer depending on the •F1 Score: 92.2%
complexity of the query. [28]. To improve the accuracy and Platform Performance
relevance of the answer, data is obtained from external
•Average Page Load Time: 1.4 seconds
databases or third-party services when required. Following
•Response Time for Chatbot Interaction: 600ms
generation of the response, it is returned to the frontend for
•Blockchain Node Sync Time: 350ms (average across GaiaNet
display by its user.Block Diagram Given in the figure 5
nodes)
demonstrates how user initiates conversation using the
interface which then through web servers reaches the engine User Feedback (from test group)
and NLP processor to generate an accurate response using the •89% of users found the chatbot interface clear and easy to use.
Database or API for maintain response accuracy and •86% reported that responses felt contextually relevant and
relevance. human-like.
•90% expressed confidence in the platform’s decentralized and
secure design.
Outcomes of the chabot is as follows:
The decentralized chatbot platform successfully met its core
objectives:
•Delivered fast, accurate, and adaptive responses through the
Phi-3 Mini 4K NLP model.
•Ensured privacy and tamper-proof data interaction using Gaia
Net’s decentralized infrastructure.
•Empowered secure and consistent client involvement through
a responsive frontend and effective backend APIs.
•Demonstrated adaptability in conversations through the
integration of user context and feedback.

Fig 5: Modular Block Diagram of Chatbot System

VII.RESULTS & OUTCOMES

Human-chatbot interface consists of an easy and efficient
communication between the user and the bot It is accessible on
website which can be used by many desktops, laptops, phone and
tablets giving user freedom of choice and convenience[29], It has
a very accessible design which is suitable for the user.Most of the
interaction happens in a text-based input and output console,
where users type in their questions and interact with the resulting Fig 6: Chatbot Response Accuracy Curve
chat bot. It retains the human-like conversation flow where
The graph in Fig 6 demonstrates Chabot accuracy as a function
messages are shown one after another in the chat window user
of user interaction over time according to the given equation. In
enabling easy conversation and responses.
the beginning, accuracy is deficient because of the cold start
When the chatbot was first released, it went through an
phase. As the bot receives more inputs, the model output
evaluation pertaining to user engagement, performance metrics,
M(x)improves. The Gaia Net Node G(n) guarantees unlink able
and overall responsiveness. The study found that the chatbot
consistency, and the personalization component P(u)
was able to handle a number of user-defined questions with
incrementally adjusts to the user. All these features combined giving them more pep, better service and quicker response times.
give rise to response accuracy R, demonstrating the system’s While these new developments are exciting, however, they don't
capabilities to evolve and improve through usage. guarantee this chatbot technology will stick around or be
universally adopted. Despite big leaps in technology, chatbots still
have a hard time thriving as an important part of regular human
interactions in all sorts of different industries, and industry experts
say it’s because they're still a bit awkward even today. The
preservation of growth and relevance requires developers to
actively address the new challenges and limitations that arise in
chatbot deployment.
Chatbots are experiencing a remarkable annual growth rate of
around 24% and are expected to be valued at $1.25 billion by 2025,
according to analysts. To achieve this anticipated potential, the
adoption of chatbot technology will be primarily dependent on its
growth beyond the current prominence of it in tech, finance and
healthcare. The most crucial issue is to enhance the adoption and
accessibility of chatbots. Why? Currently, systems fail to maintain
natural and engaging conversational flows. This is of great
concern. Several chatbots continue to follow rigid communication
patterns, leading to robotic conversations and inadequate
comprehension of user intentions. These flaws frequently hinder
users from fully relying on or using chatbot systems. To meet user
requirements and enhance interaction quality, future chatbots must
have adaptive capabilities, context-awareness, flexibility, and
Fig 7: Flow Chart of Chatbot advanced NLP models.There is an emerging trend to infuse bots
Easy No
with emotional sensitivity, in addition to enhancing smooth
Age Accuracy Precision Response
Group % % Time(ms)
UX(% of relaying of conversations. In the contemporary context, artificial
Positive) users intelligence has emancipated chatbots from merely understanding
15-20 91.5 89.6 621 87% 61
21-25 92.6 91.3 589 88% 81
signals to responding to them in a more emotionally aware manner,
26-30 92.5 91.5 589 92% 59 which has had a great impact on user experience enhancement.
Weighted
92.2 90.8 600.0 89.0 201
Chatbots can now process emotional cues and alter them to radiate
Avg
warmth and positivity which brings about trust and satisfaction.
Fig 8: Quantitative Summary of Model and Platform Metrics
These changes could be particularly beneficial for those providing
Table in Fig 8 presents a segmented analysis of key performance supportive care to patients afflicted by dementia. Emotionally
and usability metrics across three user age groups (15–20, 21–25, responsive chatbots could serve as companions, which would help
26–30) from a sample of 200 participants. The table highlights the improve the mental well-being of users by exhibiting empathy.
chatbot model’s accuracy and precision, average response times, IX CHATBOT INTEGRATIONS
and the percentage of users reporting a clear and user-friendly With the vision of better security, privacy, and security, a chatbot
experience ("Easy UX"). A weighted average row consolidates the is built on the Gaia net block chain implementing a distributed
findings, reflecting overall system performance. system. Unlike other chatbot systems that use centralized servers,
This breakdown aids in understanding demographic-specific this integration ensures that chats cannot be tampered with and
performance variations and user satisfaction trends. data cannot be lost. It's a decentralized application(utilizing smart
This is the first screen the user sees when launching the website to contracts to authenticate and validate the information while
access the chatbot and send prompts/queries. System Prompt is interacting with the participants anonymously. This compromises
used to define the default responses and behavior of the chatbot user privacy but allows conversations with the chatbot to be
according to user preference and requirement.The user can interact verified and recorded on the blockchain for infrastructure integrity.
with chatbot by typing query in the box below and using send The use of Web3. Web3. js or similar libraries helps in
button to get response to their query with the options to start a new communication over the blockchain, helping you process
chat or clear past conversations. transactions faster than ever before. Natural Language Processing
FUTURE PROSPECTS (NLP) capabilities, used in frameworks like Tenor Flow, help the
Integrating NLP into chatbot development has transformed bot understand and respond to user input.js or Open AI's API.
business practices on a big scale, especially at larger companies. The chatbot is based on an event- driven architecture where it uses
We've seen just how really effective these chatbots can be now. It's Web Sockets for real-time communication, in order to guarantee
pretty big news that NLP has made such a huge difference. For an efficient performance. This deployment option also offers
bigger businesses, NLP has really made a big splash. Larger scalability and easier maintenance if implemented as a
corporations that adopted this have gotten some pretty good results. containerized microservice (e.g., using Docker).The concept
Lots of folks are using advanced chatbots, from the mostly behind this chatbot is decentralization, which enables the users to
automatic to fully automatic systems, which help them out by have control, transparency, and trust in their digital interactions
at the core of operations. Moreover, using Natural Language 8. H.-Y. Shum, X.-d. He, and D. Li, "From Eliza to Xiao Ice: challenges and
opportunities with social chatbots," Frontiers of Information Technology &
Processing (NLP) and AI-based optimization on the blockchain
Electronic Engineering., vol. 19, no. 1, pp. 10–26, 2018.
empowers the chatbot to enhance its accuracy and contextual 9. B. A. Shawar and E. S. Atwell, "Using corpora in machine-learning chatbot
understanding of responses over time without risk to data systems," International Journal of Corpus Linguistics (IJCL), vol. 10, no.
robustness. Leveraging decentralized machine learning models, 4, pp. 489–516, 2005.
10. M. Nuruzzaman and O. K. Hussain, "A Survey on Chatbot Implementation
the system can improve and adjust its interactions while user data
in Customer Service Industry through Deep Neural Networks," in
is kept private and tamper- proof. This strategy not as it were Proceedings of the 15th IEEE International Conference on e-Business
contributes to better chatbot understanding but too fuels a more Engineering (ICEBE), 2018.
engaging, clear-cut and censorship-resistant communication 11. V. A. Bhagwat, "Deep Learning for Chatbots," Master's Projects, p. 630,
2018. doi: 10.31979/etd.9hrt-u93z.
system.
12. G. López, L. Quesada, and L. A. Guerrero, "Alexa vs. Siri vs. Cortana vs.
X.CONCLUSION Google Assistant: a comparison of speech-based natural user interfaces,"
Ultimately, working really hard on building and making International Conference on Applied Human Factors and Ergonomics.,241-
functional this NLP chatbot has been a blast that has also been 250, Springer, Cham, 2017.
13. V. K. Sarkania and V. K. Bhalla, "International Journal of Advanced
very enlightening and valuable learning experience.Through this
Research in Android Internals," vol. 3, no. 6, pp. 143–147, 2013.
project we really dug into how AI and natural language processing 14. Vichare, A. Gyani, Y. Shrikhande, and N. Rathod, "A chatbot system
can make communication much smoother and more refined in real demonstrating Intelligent Behaviour using NLP," International Journal of
time. I understand that chatbots are now much more than just Advanced Computer Science and Applications., vol. 4, no. 10, pp. 4–6,
2015.
simple conversation programs—they've truly become essential
15. N. Dandekar and S. Ghodey, "Implementation of a Chatbot using Natural
tools. Flavorful interaction is just one side of the coin. They now Language Processing," in 9th International Conference on Recent
handle tasks almost like personal helpers. Developments in Engineering, Science, Humanities, and Management, pp.
We gained profound understanding of user design principles in 99–105, 2017.
16. Kulkarni, B. Amruta, P. Savita, and K. Prof. Satish, "BANK CHAT BOT
addition to the technical knowledge of designing chatbot
– An Intelligent Assistant System Using NLP and Machine Learning," The
architectures, implementing NLP models, and managing backend journal name referenced in the citation is International Research Journal of
and frontend. It became clear to us that building automation such Engineering and Technology (IRJET)., vol. 4, no. 5, p. 4, 2017.
as a sophisticated chatbot goes beyond mere automation. 17. S. Bhalotia and S. Bisen, "Implementation of a chatbot system using AI and
NLP," The journal name referenced in the citation is International Journal
Engagement with the user must be meaningful and natural with a
of Innovative Research in Computer Science and Technology (IJIRCST).,
seamless flow of interaction. The function of sentiment analysis vol. 4, no. 1, 2018.
and its ability to humanize and soften the responses of chatbots 18. Gokaran and Ayush, "Development of Chatbot Using Deep NL P and
was particularly interesting to us. Python," no. 151263, 2018.
19. S. Raj, Building Chatbots with Python: Using Natural Language Processing
Looking to the future, we think there's an amazing chance that we
and Machine Learning, Apress, 2018.
can still make a lot of improvement. Thoughts in figuring out this 20. F. Promoteur and L. Facult, "Design and implementation of a chatbot in the
include enhancing more personal serving, really going there with context of customer support," Master thesis, Univ. Liège, 2018.
offering support in different languages and defending privacy of 21. K. Jwala, G. N. V. G. Sirisha, and G. V. P. Raju, "Developing a chatbot
using machine learning," International Journal of Recent Technology and
user data we operate with. We are really psyched about
Engineering (IJRTE)., vol. 8, no. 1, pp. 89–92, 2019.
incorporating some cutting edge ideas like blockchain for really 22. S. Ayanouz, B. A. Abdelhakim, and M. Benhmed, "A Smart Chatbot
high level security, decentralized handling of data and doing Architecture based NLP and Machine Learning for Health Care
learning directly on the device to really up the chatbot game make Assistance," ACM International Conference Proceedings. Ser., no. April,
2020.
it rock solid and trustworthy. This project has opened new doors
23. T. Spring et al., "Empathic response generation in chatbots," CEUR
for me to continue learning and invent cool stuff with AI and Workshop Proceedings, vol. 2458, Sep. 2019.
chatbots. 24. E. Pratt, "AI and Chatbots in Technical Communication – A Primer,"
REFERENCES Iiblog, p. 2, 2017.
1. D. Saha and A. Mandal, "International Journal of Computer Sciences and 25. Masi et al., "Deep Face Recognition: A Survey," Proceedings of the 31st
Engineering Open Access," Int. J. Comput. Eng., vol.3, no.1, pp.127 –135, Conference on Graphics, Patterns, and Images (SIBGRAPI), pp. 471–478,
2015. doi: 10.26438/ijcse/v7i4.184190. 2019.
2. Chung, Minjee & Ko, Eunju & Joung, Heerim & Kim, Sang Jin, 26. R. D. Sharp et al., "The Watson speech recognition engine," In Proceedings
2020,"Chatbot e-service and customer satisfaction regarding luxury brands,” of the IEEE International Conference on Acoustics, Speech, and Signal
Journal of Business Research, Elsevier, vol. 117(C), pages 587-595. Processing (ICASSP), pp. 4065–4068, 1997.
3. S. Bodrunova,"What makes users trust a chatbot for customer service? An 27. M. Solase, S. Pedgaonkar, and M. Pathade, "Technologies for Facial
exploratory interview study, "Journal of Internet Science, vol. 11193, Emotion Recognition and Chatbots for Depression Handling," Int. J.
Springer, Cham ,2018. Comput. Appl., vol. 177, no. 36, pp. 11–13, 2020.
4. J. Hill, W. Ford, I. F.-C. in H. Behavior, and undefined, "Real conversations doi:10.5120/ijca2020919841.
with artificial intelligence: A comparison between human–human online 28. L. Lin and C. J. Chen, "3D face recognition system," in Proc. IEEE Int.
conversations and human–chatbot conversations," Journal of Computers in Instrum. Meas. Technol. Conf. (I2MTC), pp. 1497–1500, 2010. doi:
Human Behavior, Elsevier, 2015. 10.1109/IMTC.2010.5488159.
5. R. Kar and R. Haldar, "Applying Chatbots to the Internet of Things: 29. R. Dsouza, S. Sahu, R. Patil, and D. R. Kalbande, "Chat with Bots
Opportunities and Architectural Elements, “International Journal of Intelligently: A Critical Review & Analysis" in Proc. Int. Conf. Adv.
Advanced Computer Science and Applications., vol. 7, no. 11, 2016. Comput., Commun. Control (ICAC3), pp. 1–6, 2019.
6. M. Mauldin, "Chatter Bots Tiny Muds and the Turing Test: Entering the doi: 10.1109/ICAC347590.2019
Loebner Prize Competition," Twelfth National Conference on Artificial
Intelligence, pp. 1621, 1994.
7. S. A. Abdul-Kader and J. Woods, "Survey on Chatbot Design Techniques
in Speech Conversation Systems,” International Journal of Advanced
Computer Science and Applications., vol. 6, no. 7, pp. 72-80, 2015.