Designing and Developing a Chatbot System for Efficient User Interaction

This research paper aims to analyze the existing manual helpdesk system and establish the challenges of the existing Chatbot systems. It proposes to design and develop a Chatbot system that allows users to get desired responses at their comforts. The chatbot system will employ inbuilt artificial intelligence to respond to queries and the answers will be correct in respect of what the user asks. The users can converse by use of any format for there is no defined format for clients to adhere.

Chatbot Research 1
CHATBOT RESEARCH
by Student’s Name:
Student’s ID
Code Course name
Professor’s name
University
City, State
Date

Chatbot Research 2
Abstract
Client interfaces for program application can come in various formats ranging from graphical,
command-line, web-application as well as voice. Common user interfaces incorporate graphical
as well as web-based programs. Periodically, the need arises for other interfaces, whether
because of multi-threaded complexity and concurrent connectivity and therefore the chatbot
based system will be able to fill the gap. The study is aimed at designing as well as developing a
chatbot system with a good user interface and with efficient responses to fulfill the requirements
of the users. The chatbot system will include functionality where the client will be capable to call
bots in a linked thread. The simulation of bots will occur by calling the callback URL that is
configured which aids bots to have communication with clients. The chatbot system uses the
algorithm of artificial intelligence to analyze the queries of the clients as well as understanding
the text of the client. The chatbot system will be a web-based application with an appropriate
user interface. The chatbot system will be more interactive such that if the queried information is
not available, one can contact admin through the portal in order to get answers. The chatbot
system will be conversing with humans to ensure they are kept updated with the ongoing
activities and any information queried is provided in time resulting to reduction of cost of
traveling to the organization and time wastage. The clients can get assistance online by use of the
web application chatbot system.

Chatbot Research 3
Table of Contents
Abstract.......................................................................................................................................................2
Introduction.................................................................................................................................................4
Chatbot that employs a limited set of rules.............................................................................................4
Chatbot that uses machine learning........................................................................................................5
Literature Review........................................................................................................................................5
Overview and Analysis of the Related Systems...........................................................................................5
ELIZA........................................................................................................................................................5
Test Weather Bot....................................................................................................................................6
Group conversation.................................................................................................................................6
Related Work...............................................................................................................................................7
Qualitative Research Method..................................................................................................................7
Proposed Research......................................................................................................................................8
Research Title..........................................................................................................................................8
Problem statement......................................................................................................................................8
The aim of the Research..........................................................................................................................8
Expected Outcome and Significance........................................................................................................8
Conclusion...................................................................................................................................................9
References.................................................................................................................................................10

Chatbot Research 4
Introduction
A chatbot refers to a computer software that has the ability to converse with human beings by
use of intelligence in texting platforms. This can be message communication, a spoken
communication as well as non-verbal conversation. Chatbot can operate on local personal
computers as well as phones although most of the times it is accessed via the internet. The
chatbot is generally viewed as an engaging program entity that human beings can converse with.
It can be inspiring, interesting as well as intriguing and it can be accessed at any place, from
earlier HTML pages to current enhanced social networking, websites as well as from quality
computers to a modern smartphone gadget [1]. Chatbot converses in various main languages,
their Natural Language Processing (NLP) professionalism from highly, poor to very bright
intelligent, funny as well as helpful..
Over the previous years, texting programs have become more common compared to networking
sites. Currently, individuals are making use of texting programs which include Facebook
Messenger, Viber, Skype, Slack, Telegram among others. This is resulting in the opening of
businesses on texting platforms which leads to extreme interaction with clients on their products.
Therefore, for interaction with many users using texting platforms, the businesses need to
embrace chatbot development that can communicate like a human. Chatbot has various
categories which include a limited set of rules as well as machine learning.
Chatbot that employs a limited set of rules
Chatbots that make use of a limited set of rules are restricted to the set rules of messaging or
commands and they are capable to react only to those messages or commands [2]. If a client
queries anything different other than the set of messages or commands that are set to the bot, it
would be able to react as expected for it is not able to understand or it lacks skills on what client
asked. These kinds of bots are good compared to the other bots.
Chatbot that uses machine learning
The chatbot that deploy machine learning functions by use of artificial intelligence. The client
does not need to be more specific while conversing with a bot for it can perceive the natural
language, not only the defined commands [3]. This type of bot grows smarter as it learns from

Chatbot Research 5
previous talks it had with individuals. The following is an example that highlights how it
operates. Here is a sample conversation between a chatbot and a human:
Human: “I need a food from your Moven Hotel.”
Bot: “Sure! Which food do you want to order?’
Human: “Hot Coffee.”
Bot: “Well! Contacting the Waiter.”
Literature Review
This literature review entails a review of the related System to chatbot system on computing
devices. The chatbot systems are not only used in firms but also in colleges to satisfy the needs
of the students without the need of traveling miles to the school [4]. The chatbot system is an
online application for responding to the queries of the user by either through a texting platform
or conversation on behalf of the company help desk. For instance, chatbot is offered by colleges
to update students on admission, activities among others.
Overview and Analysis of the Related Systems
ELIZA
ELIZA is a popular piece of a program developed in the year 1966 by Joseph Weinbaum to
emulate psychotherapist. It poses queries to humans being subjects with respect to their input and
has the capacity to carry out this without sophisticated natural language processing methods.
Alternatively, it scans what is entered for common keywords and converts it into a fresh query,
in respect with the set of rules as illustrated in its scripts. However, it is possible for a sufficiently
enhanced form of ELIZA to output the response which if sophisticated creates a natural language
processing that is being deployed and that the personal computer can literally recognize the input
it accepts [5]. At the architecture of any application of ELIZA lies the script, this script refers to
keywords, disintegration as well as assembly laws for every keyword, pre and post-processing
metonym for advanced extensibility. Theoretically, ELIZA preserves its scripts as a tree, every
keyword brings about several disintegration rules. These policies describe how an input that
resembles that keyword should be disintegrated for advanced processing [5]. Consecutively,

Chatbot Research 6
every disintegration rules progenerate various reassembly policies that state how a disintegrated
input should be recreated into a feedback.
Keywords are arguments used to establish the aim of the sentence. The essential opinion behind
this is that every sentence has got a particular target. That aim necessitates the deployment of
given words, for instance, the statement of desire obligation the deployment of words such as
want as well as need. This association goes all ways. Therefore, the deployment of given words
such as need or want, show that the client needs something. The word-intention association
permits the author of ELIZA to derive a set of efficient feedback for the correct idea if a keyword
is established.
Test Weather Bot
It is the web application that draws weather forecast for a particular town from yahoo weather
API. It inquires yahoo weather API by use of curl and gets the page with a weather predicting
data for a particular town. The forecast data are fed as JSON object. To view the weather data,
the responded JSON object is defined and retrieves the relevant data that is temperature as well
as condition. The feedback is built correctly with temperature as well as a condition of a given
place [6]. When the program is run, the weather details are shown.
Group conversation
A client who joins the chat group can begin the conversation with other group users. When a
client begins a new thread to an existing chat, many people will be alerted through email as long
as they have their email set correctly. In regard to the thread of the message, apart from the
commenter, owner of the group, individual who initiated the thread as well as everyone who has
texted on the thread except the users of the bot would be alerted through email [7]. The users of
the bot who have reacted on the thread will be alerted by clicking URL with the earlier post as an
element. This depicts how the bot communication implementation is carried out. In order to
communicate with other users, you need to create a bot account with a specific token as well as
callback URLs. For you to converse in a group, it is compulsory to state bot name with @ in the
comment. It has the ability to get bot name using normal expression and confirms whether the
needed bot is the bot client and also confirms whether it has got it in the group.

Chatbot Research 7
Related Work
Qualitative Research Method
The main aim of this research was to test the academic as well as industry literature to offer an
extensive review of quality aspect for chatbot as well as conversational agents and establish
correct quality assurance techniques. To achieve this, documentation on quality in systems of
chatbot as well as conventional agents were established via systematic studies in JSTOR, Google
Scholar as well as EBSCO Host from the year 1990 to the year 2017 [8]. The search words that
were deployed were embodied conventional agents, chatbots as well as quality assurance in
different combinations. Articles were drafted from the domains of technology, engineering,
communications, psychology as well as anthropology. The selection of researches was done in
respect of three criteria. To begin with, academic sources were emphasized and complemented
only by firm documentation from the year 2016 to the year 2017. Documentations were chosen if
they constituted at least a single search term in the head as well as abstract to guarantee the
significance of data gathering. Only documentation was chosen that possessed quality as the
main or influential element of the study. Lastly, great technical papers aimed at programming as
well as engineering components of chatbots, incorporating advancing the quality of speech
identification, were excluded.
The first search produced a sampling structure of 7,340 papers that was more refined by adding
search terms for testing, investigation, metrics as well as quality metrics. Current articles from
the year 2016 to 2017 were examined, followed by papers between 2013 as well as 2015 then
followed by the year 2007 to 2012. The choosing of time aspect was made to reduce the number
of researches for evaluation to less than 300 conference papers. 36 scholar papers were
established to be significant to the target of this article [9]. They were reinforced with 10 articles
originating from industry magazines. Past this, only 7 articles were established for the second
part of the research aiming at a quality guarantee and all were deployed in the study. We drew
quality element from every 32 papers as well as ten articles and joined them in respect of
similarity. After two or three emphases we realized that in common, they were set with the ISO
9241 factor of usability, the effectiveness, satisfaction as well as the efficiency with which
certain clients attain given goals in specified environments [10]. To be specific, effectiveness
means nearness to the correctness as well as completeness with which particular clients attain
their targets while efficiency means to how well tools are deployed to attain those objectives.

Chatbot Research 8
Proposed Research
Research Title
Chatbot system.
Problem statement
Traditionally, systems of chatbot were not known to individuals who were not concerned with
technology and with the presence of chatbot system, it is inaccurate in justifying the solution or
answer. For instance, where a college lacks the chatbot system, the students need to spend a lot
of time and resources to come to the school and bring questions to the help desk of the college
which is a cumbersome process. In such a situation the process utilizes a lot of time, resources
and can result in a communication gap between the clients and the organization or institution.
Therefore, it is necessary to develop a chatbot system that allows users to get desired responses
at their comforts.
The aim of the Research
1. To analyze the existing manual helpdesk system where you are required to avail in person
in order to pass your grievances and get the desired response which consumes a lot of
time as well as resources.
2. To establish the challenges of the existing Chatbot systems. For example, chatbot systems
with set rules cannot respond to anything outside the defined commands [11]. For
instance, syntactic indifference is the problem of these applications to understand and
include the syntactic structure of processing of the input of the user as well as
comprehension.
3. To design and develop the Chatbot system which is smarter for the efficient provision of
services.
4. To test and validate the Chatbot system.
Expected Outcome and Significance
The chatbot system will be designed by making use of an algorithm which analyzes the queries
of the clients as well as understanding the message of the user in an appropriate manner. This
system will be a web-based application. In order to respond to the questions of the users, the
clients need just to query via the chatbot that is meant for chatting [12[. The users can converse
by use of any format for there is no defined format for clients to adhere. The chatbot system will

Chatbot Research 9
employ inbuilt artificial intelligence to respond to query and the answers will be correct in
respect of what the user asks. If the answer in place is not correct or is invalid, the client is
expected to click the button with invalid answer field in the user interface in order to alert the
admin on the invalid answer then the admin can look at the incorrect answer via portal through
signing in. The chatbot system permits admin to erase the invalid answer and replace it with the
particular answer of that specific query [13]. The user can ask any activity through the system
like upcoming events implying that the client does need to travel to the institution or
organization. The chatbot system is capable of analyzing queries and react to the client just like
conversation between persons. This is aided by the use of artificial intelligence. The chatbot
system responds through constructive Graphical user display that depicts that whether a true
individual is conversing to the client. The client can question any activity through online with the
aid of a web-based application and this assists the users to be updated on the ongoing activities.
Conclusion
A Chatbot is a new software in the market as discussed above and by developing a more
enhanced chatbot system, its adoption will be supported. The proposed chatbot will make
conversations with humans easier as it is aimed at offering information as well as finishing
operations for humans they relate with. This is because the group conversation chatbot permits
the client to create an account and get a notification via email, which is an improvement. In
addition, the Test Weather Bot gives information on the weather conditions whenever the client
queries which will advance production or economic growth. I focus to advance the chatbot
system that aims at assisting individuals to carry out their work as well as interacting with
machines by use of natural language or using commands. The future chatbot system will be
capable of holding previous communications and acquire from them on what to respond to new
ones.

Chatbot Research 10
References
[1] Bowerman, G.F., International Business Machines Corp Data server administration
using a chatbot. U.S. Patent 8,250,192 , 2012. .
[2] Augello, A., Pilato, G., Machi, A. and Gaglio, S. An approach to enhance
chatbot semantic power and maintainability: experiences within the FRASI project. In 2012
IEEE Sixth International Conference on Semantic Computing (pp. 186-193). IEEE, 2012.
[3]Da Palma, W.V., Mandalia, B.D., Moore, V.S. and Nusbickel, W.L., Nuance
Communications Inc. Dialect translator for a speech application environment extended for
interactive text exchanges. U.S. Patent 8,204,182, 2012.
[4]Bradeško, L. and Mladenić, D.. A survey of chatbot systems through a loebner prize
competition. In Proceedings of Slovenian Language Technologies Society Eighth Conference of
Language Technologies (pp. 34-37), 2012.
[5]Dale, R.. The return of the chatbots. Natural Language Engineering, 22(5), pp.811-817, 2016.
[6] Duan, X.. Chatbot system and method with interactive chat log. U.S. Patent
Application 13/661,045, 2014.
[7]Rosser, R.J. and Sturges, S. Response generator for mimicking human-computer
natural language conversation. U.S. Patent 7,783,486, 2010 .
[8]Edoja, D. Method, system and program for analytics data delivering. U.S. Patent
Application 13/526,530, 2012.
[9]FitzGerald, C.W. and Shaffer, S., Cisco Technology Inc. Interactive text
communication system. U.S. Patent 7,797,387, 2010.
[10] Marciel, K.K., Saiman, L., Quittell, L.M., Dawkins, K. and Quittner, A.L.. Cell phone
intervention to improve adherence: cystic fibrosis care team, patient, and parent
perspectives. Pediatric pulmonology, 45(2), pp.157-164, 2010.
[11]Pichponreay, L., Kim, J.H., Choi, C.H., Lee, K.H. and Cho, W.S. Smart
answeringchatbot based on OCR and overgenerating transformations and ranking. In Ubiquitous
, 2016.
[12]Future Networks (ICUFN),.Eighth International Conference on (pp. 1002-1005).
IEEE, 2016.

Chatbot Research 11
[13]Takeuchi, Y. and Hoshino, A., Honda Motor Co Ltd. Chat information service system.
U.S. Patent 7,711,569 , 2010.