EDUCATIONAL CHATBOT USING CNN

1 2 3
Elizabeth Rani. G Swetha. R Pavithra. S
Department of CSBS Department of CSBS Department of CSBS
KGiSL Institute of Technology, KGiSL Institute of Technology KGiSL Institute of Technology
Coimbatore, Tamil Nadu, India Coimbatore, Tamil Nādu, India Coimbatore, Tamil Nādu, India
[email protected] [email protected] [email protected]

4 5 6
Shankari. G Deepika. S Arvind. M
Department of CSBS Department of CSBS Department of CSBS
KGiSL Institute of Technology KGiSL Institute of Technology KGiSL Institute of Technology
Coimbatore, Tamil Nādu, India Coimbatore,Tamilnadu,India Coimbatore, Tamil Nadu, India
[email protected] [email protected] [email protected]

Abstract— Marketing, supporting systems, education, health care, Natural language processing is the process of figuring out
cultural heritage, and entertainment are just a few of the fields how to turn the user's speech or text into structured data. It is
where the use of educational chatbots has grown quickly in recent utilized to provide clients with pertinent replies. One needs
years. The historical study in this paper demonstrates how the
to be extremely clear about the functionality one wants from
worldwide community's interest in chatbots has changed over time.
The benefits of chatbots in various circumstances are then
a chatbot before developing one. They are frequently created
discussed, along with the reasons for adopting them. Additionally, for commercial platforms, such as Net Banking websites, to
we emphasize how preconceptions in society affect chatbot design. handle client Q&A. Smart assistants like Google Assistant,
Following the explanation of fundamental technological concepts, Siri, Alexa, Cortana, and others are other sorts of chatbots
we classify chatbots according to a variety of criteria, including the that are developed and utilized frequently.
area of knowledge they relate to, the need they address, and other
considerations. Additionally, we highlight the key development The primary goal of a chatbot is to assist users by
environments and define the essential design of modern chatbots. providing information, answering questions, and performing
tasks without the need for human intervention. They are
Keywords— AIML-Artificial intelligence Markup Language,
widely used in customer support, e-commerce, virtual
NLP- Natural Language Process.
assistants, educational platforms, and more. Chatbots
leverage technologies like Natural Language Processing
I. INTRODUCTION (NLP) and Machine Learning (ML) to interpret user intent
and generate meaningful responses.
An application with a database, an app, and APIs to
contact other external administrations is known as an A chatbot is an artificial intelligence (AI) program
educational chatbot. Bots, however, are unable to understand designed to simulate human conversation and interact with
the customer's plans. It is a very prevalent issue that has to users through natural language. These conversational agents
be resolved. Typically, bots are educated using historical are capable of understanding text or speech inputs and
data that is solely available to them. Therefore, chatbots are providing relevant and contextually appropriate responses.
used by the majority of businesses to keep track of Chatbots can be integrated into various platforms, such as
conversation logs and analyze customer behavior. websites, messaging applications, mobile apps, and voice
These logs are used by developers to examine the assistants, to offer a seamless and interactive user
questions that customers are posing. With the help of a experience.
combination of machine learning techniques and models,
developers collaborate with their client’s inquiries and A chatbot is an artificial intelligence (AI) computer that
provide the best possible response. Compared to human mimics human conversations, including content and
training, a chatbot can be trained far more quickly and communication in natural language, using artificial
extensively. A customer service chatbot is loaded with a intelligence techniques like Natural Language Processing,
tonne of interaction logs that assist the chatbot learn what which involves speech analysis and image and video
questions to ask and what sorts of responses to offer. While processing.
regular customer care workers must follow manual
II. LITERATURE SURVEY
instructions, they are given.
The methods for developing chatbot rules have evolved,
The pattern corresponds to a set of phrases that the bots claims a study on the use of educational chatbots using deep
use, and as a result, it generates an acceptable answer for the neural networks in the customer service sector. Handwritten
clients. "Artificial Intelligence Markup Language" is the rules and templates have been the mainstay of chatbot
common structured model that is utilized to create these development strategy. These models were quickly
patterns. supplanted by end-to-end neural networks as deep learning
gained popularity. To address issues with conversational
response generation, DNN is a potent generative-based
model. With more than 70 publications on chatbots released
in the last five years, this research conducted an exhaustive
evaluation of the body of work. This analysis, which was
based on a literature review, contrasted selected studies in
terms of the method used. This study also discussed the
reasons why current chatbot models don't consider these
factors when creating responses.[1].
According to the study, a Lithuanian chatbot with intent
detection was developed using an automatic DNN hyper-
parameter. With real supervised data, optimization solved a
purpose recognition problem for the Lithuanian language.
The improvement of the NL is the primary guiding
principle.
The comprehension module is in charge of understanding
user inquiries. A Deep Neural Network classifier and a
suitably chosen word vectorization type are used to prepare
the NLU model. They studied the quick text and Bert
embeddings during their experiment [2-3].
Fig1– Flow diagram of Chatbot
Information Extraction and deep learning are two
advances that chatbots rely on, according to research on Figure-1 describes the flow diagram of the chatbot which
chatbot technology and problems. They also looked at the helps us to identify and easy to understand.
differences between transactional and conversational
chatbots. The former is classified physically to achieve a Knowing the scope and needs, such as why a chatbot is
certain goal, like booking a flight, whilst the latter is defined needed, the platform used to launch chatbots and its
manually on free-form chat logs. They have also provided a restrictions is the first stage in building a chatbot. Finding
list of commercial platforms and technologies that may be user inputs from devices and intelligence systems in the
used to build and use chatbots. finally, they have outlined form of text, speech, or image queries is the second stage.
the restrictions and potential issues with the task here [4]. Understanding the User Interface (UI) components that we
Accessible conversational user interfaces, according to can see in our programs is the third phase. There are five
research, take into account the design scope of current different types of UI elements: Form-Based Interface (FBI),
guidance and flow direction, reports, exploration, and Menu-Driven Interface (MDI), Graphical User Interface
writing on an open plan for various disability groups that (GUI), and Natural Language Interface (NLI). Building a
include clients with psychological well-being issues, mental discussion and the initial interaction come next after
imbalance, medical issues, intellectual disabilities, dyslexia, knowing the user interface components.
or learning challenges, and tangible, versatile, or ability
weaknesses. They divided this collection of directions IV SYSTEM TESTING
between those that seem to fit the available CUIs' strategy
and those that seem to provide less pressing problems and A cloud-based website and mobile testing platform called
need more research [5-7]. Browser Stack provides 3000+ actual desktop and mobile
browsers and devices for immediate, on-demand testing.
Research indicates that developing trainable chatbots that
exhibit common and human-like conversations is still a  Open a Browser Stack account. Launch the robot,
significant difficulty in artificial intelligence. Although Create a free account if you don't already have one.
Deep Reinforcement Learning appears promising for
handling this challenge, its successful implementation is still  A sample of the code is already on hand. Simply
up for debate. This research presents an innovative change and add your code.
ensemble-based methodology used for esteem-based DRL  After it has been set up, specify the prerequisites
chatbots that use constrained activity sets to represent and needed capabilities to execute the test.
importance. The training datasets in our ensemble are  If you are building your Java project with Gradle or
derived from discourse clustering, whereas in their methods Maven, adhere to the comprehensive instructions
exchange activities are derived from phrase clustering. The below. You may either configure your build path
latter intends to introduce particular agents that learn how to directly by adding the selenium jar file to your
communicate in a particular manner [8]. library to run the test case, or you can add the
maven dependency to your pom.xml file.
III. SYSTEM IMPLEMENTATION
The seven steps in the chatbot design process include
scope and requirement, input identification, UI element
comprehension, interaction design, conversation
development, and testing.
  Chatlog service is responsible for logging all the
messages in the system.
 Redis store is responsible for maintaining the state
of the messages for each user.
 Thoroughly test your chatbot to identify and fix any
issues. Evaluate its performance based on metrics
like accuracy, response time, and user satisfaction.
 Deploy your chatbot on the desired platform,
whether it's a web application, mobile app, or
messaging platform like Slack or Facebook
Messenger.

Fig 2– Testing of Chatbot.

Figure-2 describes the testing method of the chatbot. How to
server control the chatbot and how the data are handled in
the server.

Fig 3– System architecture of Chatbot.

WORKFLOW

 API Gateway passes on the incoming request from

different clients to respective services.
 The chatbot server takes in chat messages from
different clients and responds.
 The NLU engine is responsible for predicting the
intents and entities from a given text.
 The Chatbot server also acts as an action server that
calls external APIs to perform a specific action.
Fig 4 – Implementation of chatbot.
Figure 4 describes to start by identifying the purpose and
scope of your chatbot. Determine the specific tasks or
conversations it will handle. This will help you understand
what functionalities and capabilities the chatbot needs.
Decide whether you want to build your chatbot from scratch
using a programming language or use existing chatbot
development frameworks like Dialogflow, Microsoft Bot
Framework, Rasa, or IBM Watson Assistant. Each platform
has its pros and cons, so choose one that aligns with your
requirements.

Gather the necessary data for training your chatbot. This

data could include conversations, user inputs, and
corresponding responses. Preprocess the data to remove Fig 6 – Chatbot structure.
noise and format it appropriately for training. NLP is a
crucial component of chatbots as it allows them to Figure 6 describes the required chatbot, here we will be
understand and process user inputs. You can use NLP answering some queries regarding our knowledge or we can
libraries like spaCy, NLTK, or Hugging Face Transformers know about education information. Modern, tech-savvy
to perform tasks like tokenization, named entity recognition, consumers are constantly searching for the best, most
and sentiment analysis. individualized customer experiences. Meeting the avalanche
of always-changing demands can appear to be an
insurmountable feat. A chatbot, on the other hand, is one
solution that is guaranteed to please the modern customer.
Your business can quickly provide excellent service and
IV. RESULT AND DISCUSSION handle disputes for a big number of consumers at once with
the help of a chatbot. 90% of consumers, according to
Microsoft, expect customer service to be available online.
The need for AI-powered chatbots will only increase as they
become an increasingly important part of corporate
evolution. Deploying a chatbot solution right away will
prevent your business from falling behind.

Fig 5 – Website of Chatbot.

Figure 5 describes the webpage of the chatbot which is

named Dear Professor. Students can sign up for their courses
here. Their fundamental information will be gathered with
the aid of the chatbot. To replicate a conversation or
engagement with a real person, chatbots, usually referred to
as "conversational agents," are software programs that
mimic spoken or written human discourse. Visitors can
access chatbots mostly through web-based applications or
standalone apps.
 October 2018. pp. 249–258. Association for Computing
Fig 7 – Design of Chatbot. Machinery, New York, NY, USA (2018

Fig 7 describes the chatbot design and here you can see the
[2] Baldauf, M., Bösch, R., Frei, C., Hautle, F., Jenny, M.:
questions asked by the penguin and we are answering their
Exploring requirements and opportunities of conversational user
queries. According to me, chatbots and other intelligent
digital assistants are fundamentally altering the corporate interfaces for the cognitively impaired. In: Proceedings of the
landscape. Numerous chatbot-building platforms are 20th International Conference on Human-Computer Interaction
accessible to a variety of businesses, including those in e- with Mobile Devices and Services Adjunct, Barcelona, Spain,
commerce, retail, banking, leisure, travel, healthcare, and September 2018. pp. 119–126. Association for Computing
other sectors. On messaging apps, chatbots may reach a
Machinery, New York, NY, USA (2018).
larger audience and are more efficient than people. They
might soon become an effective tool for acquiring
information. [3] Beaunoyer, E., Dupéré, S., Guitton, M.J.: COVID-19 and digital
inequalities: Reciprocal impacts and mitigation strategies.
Comput Human Behav. 111 (2020).
CONCLUSION
Users can quickly enter their inquiry in natural language
[4] Bickmore, T.W., Picard, R.W.: Establishing and Maintaining
and obtain information using an educational chatbot, which
is one of the easy ways to convey data from a computer Long-term Human-computer Relationships. ACM Trans.
without having to think of appropriate keywords to look up Comput.-Hum. Interact. 12, 293–327 (2005).
in a search or explore multiple web pages to collect
information. The design and execution of the chatbot have
[5] Bigham, J.P., Aller, M.B., Brudvik, J.T., Leung, J.O.,
been discussed in this paper. According to the study above,
Yazzolino, L.A., Ladner, R.E.: Inspiring blind high school
the many methodologies and approaches used to construct
chatbots cause the growth and progress of chatbot design to students to pursue computer science with instant messaging
proceed at an unpredictable rate. A chatbot is a fantastic tool chatbots. SIGCSE Bulletin. 40, 449–453 (2008).
for speedy user contact. They assist us by amusing us,
saving us time, and responding to queries that are difficult to [6] Boshoff, A.: Accessibility and chatbots: how to make your
locate.
chatbot user-friendly for everyone. Roger wilco.
Chatbots provide a natural and user-friendly way for https://www.rogerwilco.co.za/blog/accessibility-and-chatbots-
people to interact with technology. They leverage natural how-make-your-chatbot-user-friendly-everyone (2018)
language processing and understanding to comprehend user Accessed 17 December 2020.
inputs and generate relevant responses. Unlike human
agents, chatbots can operate 24/7 without fatigue or the need
[7] Brave, S., Nass, C., Hutchinson, K.: Computers that care:
for breaks. This constant availability allows businesses to
engage with customers at any time, significantly improving investigating the effects of orientation of emotion exhibited by
customer service. an embodied computer agent. International Journal of Human-
Computer Studies. 62, 161–178 (2005).
Chatbots have become an essential part of modern
communication and business strategies. They offer
.
numerous benefits, including improved customer support,
increased efficiency, and enhanced user experiences. As
technology continues to advance, chatbots are becoming
more sophisticated, capable of understanding and
responding to human language with greater accuracy.

REFERENCE

[1] Abdulrahman, A., Kuber, R., Branham, S.M.: “Siri Talks at

You”: An Empirical Investigation of Voice-Activated Personal
Assistant (VAPA) Usage by Individuals Who Are Blind. In:
Proceedings of the 20th International ACM SIGACCESS
Conference on Computers and Accessibility, Galway, Ireland,