Journal of Advancement in Software Engineering and Testing

Volume 3 Issue 1

Research Paper on Chatbot for Student Admission Enquiry

Punith S1, Chaitra B2, Veeranna Kotagi3*, Chethana R M4
1,2
Students,3,4Assistant Professor
Department of CSE, East West College of Engineering, Bengaluru, India.
*Corresponding Author
E-mail Id:[email protected]

ABSTRACT
In this research paper, "Chatbot for Student Admission Enquiry" we describe about the
software/tools used in it. We are using Rasa NLP, we have used dialogflow to make
interaction with chatbot and also the NLP toolkit. This chatbot was developed using Machine
Learning Algorithms. Chatbots are prepared to change UI design. Chatbots are programs
that duplicate human conversation using Artificial Intelligence (AI). It is proposed to be an
authoritative modest assistant, delight reason, helping one to complete tasks running from
tending to questions, getting driving heading, turning up the indoor controller in sharp home,
to playing one's favored tunes etc. Chatbots are exhibited on various techniques, for instance,
data base, AI based. Artificial intelligence based chatbots yields progressively reasonable
results. Chatbot which gives responses reliant on the setting of conversation will when all is
said in done be all the more simple to utilize. One of the noteworthy goals in the field of
Human Computer Interaction (HCI) is the diagram of common and instinctive affiliation
modalities. In any case, chatbots are at present being grasped at a high rate on PC talk
stages. The advancement at the focal point of the climb of the discussion bot is ordinary
language processing. Most business chatbots are dependent on stages made by the
development mammoths for their standard language dealing with. These fuse Amazon Lex,
Microsoft Cognitive Services, Google Cloud Natural Language API, Facebook DeepText,
and IBM Watson. Stages where chatbots are passed on join Facebook Messenger, Skype, and
Slack, among various others.

Keywords:-Natural language processing, chatbot, Artificial Intelligence (AI) etc.

INTRODUCTION everything is right now getting related

A chatbot is an instance of emotional with web. It is extremely simple to utilize
figuring system that mirrors human approach to manage benefit everything to
conversations to give instructive, esteem doorstep. There are different sorts of
based, and conversational organizations. customer help open like live talk reinforce
Despite that no matter how you look at it organization, phone (telephone)
gathering, chatbots still experience the evil organizations. In any case, for all such
impacts of different execution issue as a assistance organizations given by human
result of imprisonments with their to human puts aside some push to respond
programming and training. Chatbots must to customer’s question. As the amount of
destruction different issues, including clients develops the holding time increases
flexibility, low-dormancy, and privacy. as well, which realizes poor client
Today time is having numerous electronic satisfaction. One of the critical goals in the
organizations like E-business, field of Human Computer Interaction
Entertainment, Virtual assistance and some (HCI) is the system of run of the mill and
more. There is radical augmentation in the natural affiliation modalities. Specifically,
domain of web organization, where different endeavors have been centered

HBRP Publication Page 1-9 2020. All Rights Reserved Page 1

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

around the improvement of structures to customer has as of late sent. It could be a

talk with the client in a trademark welcome, conversation subject, or even an
language. image. Chatbots are normally used in
return systems for various sensible
The world is moving towards the purposes including customer help or
automation of human effort with the information acquisition. Most central
help of machine learning, artificial chatbots work by organizing a customer’s
intelligence and robotics. While there is commitment with a predefined set of talk.
a huge scope of improvement, some of For example, a customer saying "Thank
the key areas that can be targeted for you" will result in the chatbot saying
automation are responder system in "You’re Welcome".
education, healthcare etc. An artificial
intelligence based bot can replace Chatbots continue creating in
human efforts without compromising on pervasiveness with 80% of associations
the quality of response. Machine planning to use one by 2020.
learning and artificial intelligence is Notwithstanding the way that it may feel
used to implement chatbot along with like the term 'chatbot' has starting late
python scripts. entered the open word reference, they truly
have a more drawn out history than you
Past chatbots, Conversational AI suggests may foresee.
the usage of illuminating applications, talk
based partners and chatbots to automate LITERATURE SURVEY
correspondence and make modified "Undertaking Crowd Computing for
customer experiences at scale. The term Human Aided Chatbots" The dream of a
"ChatterBot" was at first initiated by human-like especially sharp PC associate
Michael Mauldin (producer of the first isn't new. In science fiction books and
Verbot, Julia) in 1994 to depict these movies, from Hal in “2018: A space
conversational undertakings. Today, most odyssey” (1968) to Jarvis in “Iron Man”
chatbots are found a good pace of remote (2018), singular assistants helped legends
partners, for instance, Google Assistant for an amazing duration and to manage
and Amazon Alexa, by methods for their work commitments. In investigation,
illuminating applications, for instance, the primary conversational pros appeared
Facebook Messenger or WeChat, or during the 60's. They were unequivocally
through individual affiliations' applications modified, using rule-based philosophies.
and locales. Chatbots can be organized Chatbots are content based conversational
into usage classes, for instance, administrators, living in task individual
conversational business (online business applications (for instance Facebook
by methods for talk), assessment, Messenger, Telegram, Whatsapp and
correspondence, customer help, structure, WeChat), and duplicating a conversation
creator gadgets, preparing, beguilement, with a human to give enlightening, esteem
finance, sustenance, games, prosperity, based, or conversational organizations.
HR, publicizing, news, singular, Given a customer request, the language
productivity, shopping, social, sports, getting (LU) fragment determine the
travel and utilities. Most of the chatbots customer's point and the related
are furnished with a task individual sort information; the Action Execution and
interface with a commitment from a Information Recuperation (AEIR) part
customer and a yield from the chatbot. The plays out the referenced exercises, or
chatbot structures the customer’s data and recoups the information of excitement
yields an answer reliant on what the from its data source; the response age

HBRP Publication Page 1-9 2020. All Rights Reserved Page 2

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

(RG) portion develops a response to the for making the Assistant (by Speaktoit), a
customer, finally the trade, the board (DM) virtual mate for Android, iOS, and
section jam and update the setting of a Windows Phone PDAs that performs tasks
conversation to request missing and answers customers' request in a
information, to process clarifications by trademark language. Speaktoit has
customers, and to ask follow-up requests. furthermore made a trademark language
Top tier chatbots are yet far from being getting ready engine that joins
extraordinary, fighting to function conversation setting like trade history,
admirably for customer requests and to zone and customer tendencies.
carry on a huge conversation. All of the
chatbot's fragments experience the evil In May 2012, Speaktoit got an undertaking
impacts of different limitations when used round (financing terms undisclosed) from
in obvious circumstances. For instance, the Intel Capital. In July 2014, Speaktoit shut
LU may bomb translating customer their Series B financing drove by Motorola
requests, the DM may miss clarification Solutions Venture Capital with
requests for missing information, the participation from new theorist Plug and
AEIR presumably won't find the Play Ventures and existing benefactors
referenced information or not execute the Intel Capital and Alpine Technology Fund.
privilege referenced movement, and the
RG likely won't give a decent response. In September 2014, Speaktoit released
api.ai (the voice-engaging engine that
PROBLEM STATEMENT powers Assistant) to outcast creators,
Artificial intelligence chatbot is a allowing the extension of voice interfaces
technology that makes interactions to applications subject to Android, iOS,
between man and machines using natural HTML5, and Cordova. The SDK's contain
language processing. We are building a voice affirmation, ordinary language
chatbot for the college which answers the understanding, and substance to-talk. api.ai
user queries regarding the information offers a web interface to develop and test
about the college. The bot is integrated to conversation circumstances. The stage
the college website. The bot replies to the relies upon the typical language getting
user with appropriate answers within a ready engine worked by Speaktoit for its
fraction of second. It also contains Assistant application.Api.ai grants Internet
analytics part which displays the number of Things developers to fuse normal
of users who visited the website and used language voice interfaces in their things.
to bot to get the information about the Partner and Speaktoit’s locales by and by
college. This analytics part can be seen redirect to api.ai’s site, which derails the
only by the admin. Humans often get Dialogflow site.
bored answering the same question again
and again and more over human can Google obtained the association in
handle at most 2-3 users but by September 2016 and was from the outset
implementing this bot, it can handle n known as API.AI; it offers instruments to
number of users and it is available all time engineers building applications
in a day. ("Actions") for the Google Assistant
remote aide. It was renamed on 10 October
RESULTS & DISCUSSION 2017 as Dialogflow. The affiliation halted
Dialogflow is a Google-asserted specialist the Assistant application on December 15,
of human–PC affiliation advancements 2016.
reliant on regular language conversations.
The association is generally mainstream Voice and conversational interfaces made

HBRP Publication Page 1-9 2020. All Rights Reserved Page 3

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

with Dialogflow works with a wide extent that are critical to entertainment and
of contraptions including phones, settlement adventures. Dialogflow in like
wearables, vehicles, speakers and other manner joins an examination gadget that
splendid devices. It supports 14+ tongues can check the dedication or session
including Brazilian Portuguese, Chinese, estimations like use plans, idleness issues,
English, Dutch, French, German, Italian, etc joins an examination gadget that can
Japanese, Korean, Portuguese, Russian, evaluate the dedication or session
Spanish and Ukrainian. Dialogflow estimations like use structures, inertia
supports an assortment of organizations issues, etc.

Fig.1:-Outlook of dialogflow

HBRP Publication Page 1-9 2020. All Rights Reserved Page 4

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

Dialogflow Architecture

Fig.2:-Dialogflow Architecture
There are 4 ideas that must be client.
comprehended to get a decent image of There is part of inquiries on gatherings on
Dialogflow's engineering. the most proficient method to send ‘pop-
up message’ from Dialogflow specialists
Solicitation Reaction to the end client. (For example send them
You can condense a Dialogflow specialist an update by means of Facebook envoy
as a progression of (demand, reaction) sets dependent on a past discussion). The
i.e. solicitation first and reaction next
worldview implies that this component
1 User sends a solicitation to the isn’t bolstered in Dialogflow and the best
Dialogflow operator. way to accomplish pop-up messages is by
2 Some NLU enchantment occurs. sending ‘out of band’ messages.
3 Dialogflow sends a reaction back to the

Fig.3:-Request-Response
Intent Mapping grasped is reason mapping. At the point
The second thought that ought to be when your Dialogflow administrator gets

HBRP Publication Page 1-9 2020. All Rights Reserved Page 5

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

the customer’s requesting, it will plan.

endeavour to layout to an overview of pre- 5. Dialogflow will parse the response,
portrayed points. Objectives are made by separate pertinent information and use
the bot fashioner as a strategy for telling it in the response sent to the customer.
the Dialogflow organization ‘here is a
once-over of assignments that the Compromises
customer may ask the chatbot to do’. Despite the features above, Dialogflow in
like manner revealed an API. This API
To perceive how the desire mapping truly licenses you to naturally send sales to your
works, you need to see some reinforcing Dialogflow masters. The sales will be
thoughts-components, settings, parameters, mapped to an arrangement and a fitting
etc. You can make a gander at my walk by response will be sent to the API visitor,
step manual for help you with this. comparably as if the requesting message
was made into Dialogflow’s solace. This
After Dialogflow maps the customer's infers you can ‘introduce’ a Dialogflow
requesting to one of the pre-described pro into your present applications and
plans, it will send a response (which is organizations.
moreover shown by the bot creator) back
to the customer. Gathering Everything
Let us consider Dialogflow’s worked in
Fulfilment web demo mix. In case you should see it,
Once Dialogflow maps the customer’s all things considered, take a gander at this
requesting to one of the pre-described chatbot which offers direction on SEO.
points, it might need to play out some This is using the certain web demo blend
business method of reasoning before gave by Dialogflow. Regardless, the web
sending a response. Instead of making demo compromise has a huge amount of
‘business basis LEGO squares’ (which requirements (for instance can’t show
some other bot frameworks do), the organized substance, line breaks, pictures,
Dialogflow individuals decided to use a etc.).
thought called Fulfilment.
Here is the way by which it works: So you can consolidate your site with a
1. Dialogflow fathom the customer’s Dialogflow chatbot using its REST API,
sales and concentrates sorted out and get an even more ostensibly
information from their message. captivating chatbot. For example, my
2. The composed information is sent to SupportBot reinforces pictures, content
some code running on some web have orchestrating, line breaks, gets, list
which is kept up by the bot producer - selectors, etc.
this code is known as a webhook .
3. The webhook plays out the significant Exactly when the customer partners with
business method of reasoning. my SupportBot, this is the way by which
4. It will re-establish a response to your the data streams:
Dialogflow administrator in a specific

HBRP Publication Page 1-9 2020. All Rights Reserved Page 6

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

Fig.4:-Data Flow

Fig.5:-Session flow

Float your cursor over the desire names composed to.

(blue boxes) to see the going with  The level of customers that left the
information: point.
 The level of all customers that
organized the point. Snap Explore to see the graph in a greater
 The number of sales the point was arrangement. While seeing the outline at

HBRP Publication Page 1-9 2020. All Rights Reserved Page 7

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

the present time, can tap on a desire to International Journal of Engineering

broaden the diagram and show which and Social Science. ISSN: 2249-
intentions were facilitated straightaway. 9482. 2012.2(5) (May 2012):1-6p.
4. Sandhya Soman, Piyush Kumar
CONCLUSION Pareek 2020 J. Phys.: Conf. Ser. 1427
Our bot will act as a person at enquiry 012010
office whom they can ask any questions 5. PK Pareek. Optimization of the
regarding admission and college critical parameters affecting the fuel
environment at any time in more lid performance 2019-28-2413
interactive way. Thus, reducing workload 6. Piyush Kumar Pareek. ROC Structure
of the admission cell personnel. This Analysis of Lean Software
system can be integrated on college Development in SME’s Using
website and thus students who desire to get Mathematical CHAID Model (May
admission in the college can clear their 17, 2019)
doubts from their home. We have provided 7. KD Menon, A Raj Jain, Piyush Kumar
a feedback mechanism through which user Pareek. Quantitative Analysis of
can rate bot’s responses and thus bot can Student Data Mining.2019.
respond more accurately next time. 8. S HS, S Soman, Dr.Piyush Kumar
Pareek. Fast and Efficient Parallel
Dialogflow is a google product which Alignment Model for Aligning both
helps in creating a chatbot. Dialogflow is a Long and Short Sentences. 2019
specifically designed for creating chatbot 9. MR Suresh, PC Behera, C Dash, PK
and it is considered to be one of the best Pareek. Face Recognition with
platforms. In dialogflow the bot is trained Convolutional Neural Network.
with various intents and respective Journal of Advance Research in
responses are added. This model can be Mobile Computing.1 (1)
updated as and when new rule is 10. A Pai, VS Veesam, BS Babu, Piyush
implemented in the college. This bot can Kumar Pareek. Six sigma approaches
maintain context and can give responses used in implementing in supply chain
based on context. management: A Review. Research and
Applications of Web Development
REFERENCES and Design 1 (2):12-16p.
1. Pareek, P.K., Nandikolmath, T.V., 11. S Vikas, GV Attimarad, P Pareek.
Gowda, P. FMEA Implementation in a Implementation of a Low-Cost and
Foundry in Bangalore to Improve Non-invasive System for the
Quality and Reliability. International Measurement and Detection of Faulty
Journal of Mechanical Engineering Streetlights. Journal of Advancement
and Robotics Research.2012.1(2):81- in Electronics Design 1 (1), 19-30p.
87p. 12. Piyush Kumar Pareek. Questionnaire
2. Nandikolmath, T., Pareek, P. K., & Survey using CHI-Square Test in Six
SA, V. K. Implementation of a Lean Sigma in SME’s in Bengaluru.
model for carrying out value stream Advancement in Image Processing
mapping in manufacturing industry. and Pattern Recognition 1 (1):20-24p
International Journal of Mechanical 13. Piyush Kumar Pareek, Analysing
Engineering and Robotics Tools in Six Sigma in SME’s in
Research.2012.1(2):88-95p. Bengaluru, Journal of Advancement in
3. Piyush Kumar Pareek, Praveen Software Engineering and Testing 1
Gowda et al Ergonomics in a Foundry (1), 19-29
in Bangalore to improve productivity. 14. Piyush Kumar Pareek. Six sigma

HBRP Publication Page 1-9 2020. All Rights Reserved Page 8

 Journal of Advancement in Software Engineering and Testing
Volume 3 Issue 1

approaches used in implementing in Automated Approach Used in Lean

supply chain management: A Review. Six Sigma While Implementing in
Journal of Advancement in Software Supply Chain Management: A
Engineering and Testing 1 (1), 14-18p Review, Research and Applications of
15. A Pai, VS Veesam, Piyush Kumar Web Development and Design.
Pareek, BS Babu. Challenges in 1(1):7-13p
SME’s ANOVA ANALYSIS PART-2 in 19. D Chandramma Piyush Kumar
Bengaluru. Research and Reviews: Pareek, ANOVA Analysis Part One of
Advancement in Robotics. 1(1):9-15p. Challenges in SME’S in Bengaluru,
16. A Pai, VS Veesam, BS Babu, Piyush Research and Reviews: Fast and
Kumar Pareek, ANOVA Analysis Part Accurate Parts of Speech Tagging for
One of Challenges in SME’S in Kannada-Telugu Pair, International
Bengaluru, Research and Reviews: Journal of Applied Engineering
Advancement in Robotics 1(1):1-8p. Research. 13(10):7857-7867p
17. A Pai, VS Veesam, BS Babu, Piyush 20. Piyush Kumar Pareek. ANOVA
Kumar Pareek. ANOVA Analysis Part Analysis Part One of Challenges in
One of Challenges in SME’S in SME’S in Bengaluru, Research and
Bengaluru, Research and Reviews: Reviews: K Swathi, P Shetteppanavar,
Understanding the Adaptability of An efficient machine translation
SME’S in Bengaluru, Research and model for Dravidian language , 2017
Reviews: Advancement in Robotics 1 2nd IEEE International Conference on
(1), 16-22 Recent Trends in Electronics
18. A Pai, VS Veesam, BS Babu, Piyush 21. PK Pareek. An adoptive Model for
Kumar Pareek. ANOVA Analysis Part lean software development in small
One of Challenges in SME’S in and medium level firms in Bengaluru.
Bengaluru, Research and Reviews: 2016
Comparative Analysis of
Conventional Approach and

HBRP Publication Page 1-9 2020. All Rights Reserved Page 9