You are on page 1of 4

AI and MI Based Chatbot for Student Database

and Admission
Praveen Kumar R
Assistant Professor
Department of Computer Science and Engineering,
Cambridge Institute of Technology, Bangalore, India
praveenkumar.cse@cambridge.edu.in

Mohammad Hussain Sheriyar Shaik


Student Student
Department of Computer Science and Engineering, Department of Computer Science and Engineering,
Cambridge Institute of Technology, Bangalore, India Cambridge Institute of Technology, Bangalore,
hussain.17cs0083@cambridge.edu.in India sheriyar.17cs148@cambridge.edu.in

SVS Sumanth
Student
Department of Computer Science and Engineering,
Cambridge Institute of Technology, Bangalore, India
sumanth.17cs163@cambridge.edu.in

Abstract— Many people adopt sensible Assistant Devices like on Dialogflow, Actions on Google and base Realtime
Google Home. Currently, days of only partaking with a service information. the educational organizations and different school
through a keyboard area unit are over. The new modes of user industries were among the primary Industries to adopt new
interaction area unit motor-assisted partially by this analysis technologies.
can investigate however advancements in computer science and This integration has grownup massively, serving to
Machine Learning technology area unit being employed to organizations reach a wider client base enabling them to
enhance several services. especially, it'll consider the event of perform their handily academic organizations have become
google assistants as a channel for data distribution. This project ever additional competitive with an associate degree other to
is aimed to implement an Associate in Nursing android-based adopt the most recent advancements in technology to supply
chatbot to help with Organization basic processes, exploitation
an improved delivery service to satisfy users.
of google tools like Dialogflow that uses language process
information processing, Actions on Google and Google Cloud The Organizations area unit currently enables the utilization
Platform that expose computer science and Machine Learning of technology thus, customers will perform additional tasks
strategies like language understanding. permitting users to online victimization the voice-based mostly and informal like
move with the google assistant exploitation language as input obtaining results, attendance, personal and educational data
and to coach the chatbot i.e. google assistant exploitation remotely and intelligent chatbots to extend client service and
Dialogflow Machine learning tool thus me and a few acceptable assist staff & users. Generally, chatbots area unit an easy
strategies so it'll be able to generate dynamic responses. The software package programs which will answer client prompts.
chatbot can enable users to look at all their personal
the main focus of this project is to implement these new
educational data, schedule conferences with higher officers, and
automate the organization method and organization resources technologies to make an associate degree intelligent chatbot
data all from intervals the chatbot i.e. Google Assistant. This Google assistant to change organizations to charm millennian
project uses OAuth authentication for security purposes. and doubtless gain a long user. The planned system takes an
Dialogflow helps to grasp the user’s question by exploiting academic establishment as a reference.
machine learning algorithms. by exploiting this google assistant
we have a tendency to area unit attending to use the Cloud II. METHODOLOGY
Vision API for advancement. we are going to use Dialogflow as
the key half to developing Google assistant.
This planned system is employed to make a google assistant
Keywords Google Cloud Platform, Natural Language Processing, simulate somebody's oral communication to help users with
Actions on Google, Smart Assistants; not solely academic wants it may be applied to any
organization and to supply a lot of personal expertise. In
I. INTRODUCTION today’s world, there are such a lot of advancements in
computer science, machine learning techniques, improved
Google Assistant helps users to move with our product and power for higher cognitive processes, and larger availableness
services in any organization victimization in variety of text- of domains and corpus, which have enhanced the utility of
based and informal interfaces like voice by victimization the group action a conversation larva into applications we will
Dialogflow tool, that have the integral machine learning integrate the chatbot. Users are able to raise any establishment
algorithms perceive to know to grasp the users promptly and connected queries in language that they're snug mistreatment
supported the users question our chatbot can understand the like reading account info, grades, transactions and check
coaching phrases and generates the dynamic response to the balance, Subjects listed etc. details the mistreatment of the
user. during this paper, we tend to area unit attending to study assistant.
The chatbot can determine Associate in Nursing perceive
what the user is asking and generate an acceptable response
RESULTS AND DISCUSSION
supported by the informal context. Immediate responses are
provided by the chatbot to redeem the requirement for the user
to own to decision or visit their faculty branch Associate in
Nursing and wait in queue so as to induce through to an
authority for help. so as to create the application safer O auth
Authentication protocols are integrated and provide high
security making certain solely registered users will gain access
to their account preventing the chance of fraud.
We will use the Google Cloud APIs for this project and we
can develop many alternative options. This system uses a base
of operations Real-time information we will build wealthy,
cooperative applications by permitting secure access to the Fig 4.1 college website
information directly from client-side code. knowledge is
persisted regionally, and even whereas offline, real-time events 4.1 Student
still fireplace, giving the top user responsive expertise. This The student modules will facilitate the scholars for accessing
system additionally uses the advanced technologies like their tutorial information over the voice and informal based
Google Cloud Platform, Dialog flow, Actions on Google and mostly system, for instance, once a user says “I need to grasp
Google Cloud vision API i.e. Cloud Vision API permits my grades or result “It can generate a response supported by
developers to simply integrate detection options inside the trained phrases and uses the machine learning methods.
applications, together with image labelling, face and landmark These modular going to be shown on the user logins with
detection, optical character recognition (OCR), and tagging of faculty the school the faculty UserId and college MailId like
specific content. Google Cloud automotive vehicle millilitre an in the similar method it gets do all the fundamental
Vision API allows you to form a custom machine learning method.
model for image labelling.
The system is joined with Google Assistant to supply
marvellous expertise to users, this project is accessible to each
user wherever ever the google is there. The users of this
project aren't ought to install the app on their device, solely
users ought to decide on the project by exploiting the
invocations like “Talk to CIT bot”, once we decide on the
invocation this app asks to you sign up for app exploitation
varied ways like sign up with Google, Facebook, LinkedIn and
manual username and word technique.

III. SYSTEM ARCHITECTURE

The system architecture consists of JavaScript programming


language, Actions on Google, Dialogflow, and Firebase
Realtime Database. Dialogflow which is an integrated
development environment from Google is used to develop
chatbots using various programming languages.

Fig 4.2 Welcome Page

Fig 3.1 System Architecture


Fig 4.4 Respond page

Fig 4.3 Asking Query Then it'll be asked you to link your account After productive
authentication, it'll be redirected to the app and it'll say
The basic method of this project is shown the below diagrams greetings to the user and asks however am I able to assist you
thus, we are able to visualize the operating method of the with organization resources.
project
V. WORKING
4.3. Google Cloud Vision API Then the user can move with this google assistant project
Google Cloud Vision API is a Pre-Trained Machine Learning and as an example, if a school asks like this “I ought to get
model that helps to derive insights from pictures and videos. student details “, then the app asks you to enter the register
range of the coed when coming into it'll fetch and provides the
dynamic response to the user. Like this in a similar approach,
it will do all tasks.
Whenever we say over voice to the app i.e. Dialogflow will
convert the speech into text and understand and compare the
trained phrases with help of Intents and Agents and if in case
we enable the webhook the query will be sent to the database,
here I have used Firebase Realtime Database will send a
response to the user by converting the text into speech by the
Google Assistant. With help of Google Cloud Vision API,
we can detect the faces of students and take any organization
to next level of advanced technology.
For using the Google API library scopes, we need to
Fig4.5 To check the college website https://console.cloud.google.com/apis/credentials?projectProj
ect name, we need to fill the Consent screen for verification
You can get insights including image labelling, face detecting
by the Google Security team, if they approve only we need to
and landmark detection, optical character recognition (OCR),
use their scopes such as Scheduling meetings with some
and tagging of explicit content. In this project, we can detect
libraries, Contacts i.e. People API and there are a lot of API’s
the faces of students and take their details within seconds of
are available to use. With that API we can create a super AI
processing the data. Open Android mobile or Google home or
& Machine Learning Model and do a lot more with
good devices that have google assistant i.e. Open google
technology. For approving our Google Assistant project, we
assistant and simply say “Talk to CIT bot”.
also need to host a Privacy Policy for that App in our own
Domain.
Like I have hosted it on my domain
https://google.technology/PrivacyPolicy.html. In Google
Cloud Platform we can manage all the application insights
such as API requests, responses and errors etc. In Actions on
Google Console we can test the app in a simulator over the
voice and text-based and also we can test the app in a simulator
mobile before publishing it. For that, we need to Sign In with
the same account on our mobile for Google Assistant.

Fig 5.3 Dialog flow Fulfillment & Webhook page

VI. CONCLUSION

Therefore, during this approach, we are able to produce a


synthetic Intelligence & Machine Learning primarily based on
google assistant for any organization. This method is simpler
when put next to alternative procedures. which will alter the
Fig 5.1 Project Overview Details
essential and sophisticated tasks and reduces the worker
intervention of the organization to unravel the client queries.
because of the higher than expressed methodology, we are able
to improve the Machine Learning techniques in Dialogflow
and base period of time information provides numerous
options and with OAuth authentication, our application
becomes safer and responsive and Google cloud or the other
cloud services we are able to use like Microsoft Azure, Google
Cloud etc.

REFERENCES
1. https://dialogflow.com/docs/reference/agent/intents
2. https://cloud.google.com/docs/
3. https://firebase.google.com/docs/database/security
4. https://firebase.google.com/docs/database/security/indexing-data
5. https://developers.google.com/assistant
6. Sristhi Chawla (2018, May 29). Introduction to Natural Language
Processing:
7. Kshitiz. V. Wadanka et al., "Chatbot:An Application of AI",
International Journal of Research in Engineering, Science and
Management, Vol.1 Issue 8, September 2018, pp. 139 - 141.
8. Razwan Pascanu et al. "On the difficulty of training recurrent neural
networks" International Conference on Machine Learning, Atlanta,
Georgia, USA, 2013.
9. Azvine B., Djian D., Tsui Ching K., Wobcke W (n.d). Retrieved from
http://www.cse.unsw.edu.au/~wobcke/papers/ia.pdf
10. Chatbot Learning: Everything you need to know about machine
learning chatbots (2020). Retrieved from
https://www.whoson.com/chatbots- ai/chatbot-learning-everything-
need-knowmachine-learning-chatbots/
11. https://chatbotsmagazine.com/the-ultimateguide-to-designing-a-
chatbot-tech-stack-333eceb431da

Fig 5.2 Project API & Traffic page

You might also like