See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/337400748

Ai Chatbots: Transforming the Digital World

Chapter · January 2020

DOI: 10.1007/978-3-030-32644-9_34

CITATIONS READS
23 2,174

3 authors, including:

Shweta Paliwal Vishal Bharti

DIT University Maharishi Markandeshwar University, Mullana
12 PUBLICATIONS 87 CITATIONS 65 PUBLICATIONS 256 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Vishal Bharti on 24 August 2022.

The user has requested enhancement of the downloaded file.

 CHAPTER 2
AI CHATBOTS: TRANSFORMING THE DIGITAL WORLD
Shweta Paliwal , Dr. Vishal Bharti , Dr. Amit Kumar Mishra

E-mail: shwtplwl23@gmail.com, hod.cse@dituniversity.edu.in, ak.mishra@dituniversity.edu.in

Department of Computer Science and Engineering

DIT University, Dehradun, Uttrakhand, INDIA

INTRODUCTION
In today’s era technology is booming at a breakneck speed and Artificial Intelligence is emerging
as a game changer. From the virtual assistant SIRI to the self driving cars and to autonomous
weapons AI has fascinated the concept of science fiction and is driving the world towards
automation. Artificial Intelligence is defined as the simulation of human intelligence process
(acquisition of information, reasoning) by machines. The concept of artificial intelligence coined
back in the year 1956, at Darmouth Conference organised by John McCarthy .He is known as the
‘father of artificial intelligence’ and developed the LISP programming language which later on
became an important part in machine learning. We can say that AI is study and design of
intelligent agents and these agents in today’s scenario are termed as chat bots. AI has made the
chat bots more lifelike than ever before. We have also tried to depict how the Chatbot can be
created by creating an example of the chat bot AMANDA.

Keywords: NLP, AIML, AMANDA, Python

1. CHATBOT
From ordering a pizza to scheduling a meeting we all are surrounded by robots. Chat bots have
established a well time ground between automation and the use of internet that have changed in
past few years. “Chat bots are a software program that performs cognitive service functions
along with the understanding of the natural language”. Chat bots have defined the interaction
between humans and machines in a most simplified manner. This technology started within
1960s with an aim to examine if the Chatbot system will be able to fool the real humans. With
the popular usage of private machines the needs of colloquial agents have become more intense.
To carry out the human computer interaction in the most efficient manner the users must be
allowed to express their interests or queries by speaking or typing acted as a driving force behind
the development of chat bot. It has been predicted that by 2022, almost 90% of the customer
enquiries will be dealt by these automated agents[1]. Chat bots allow business to deliver in a
personalised manner by integrating message, operations and human support in one experience.

1.1 HISTORY OF CHAT BOTS

In the year 1950, Alan Turing an English computer scientist put forward a question “Are
machines able to think” and he answered it by bringing the concept of Turing test. Turing
test is a methodology to determine if a computer can think like human which marked the
beginning of artificial intelligence
 The first Chatbot ELIZA was created in 1966 by Joseph Weizenbaum. Although she failed
in Turing test but had laid the foundational structure of chat bots which includes pre-
programmed responses and keywords.
A Chatbot named PARRY was able to simulate a person with paranoid schizophrenia. Only
about 48% were able to identify the difference between PARRY and human. It was created
by Kenneth Colby and then RACTER in 1983. Later on in 1995, ALICE a language
processing bot who can work online came into existence, but she was unable to pass the
Turing test. In 2001 smarter child was introduced.The next decade 2010 – 2015 marked the
presence of SIRI (2010), GOOGLE NOW (2012), ALEXA (2015) and CORTONA (2015).
These bots were able to respond to voice commands and are capable of performing internet
searches along with some other tasks.
In present creation of chat bots based on machine learning and neural networks can be seen in
progress. Here the figure 1 will describe how the Chatbot came into existence.

Fig.1 History of chat bots

2. ELIZA: THE FIRST CHATBOT

Eliza is a natural language conversation program created in 1966 based on objects oriented
JavaScript and was first implemented on IBM 7094. It performed best when the interaction
was carried out with the help of a typewriter. Eliza functions by recognising the keywords
from the input to breed a response from pre programmed responses, hence creating an
illusion of interaction with a human being though the process is a mechanised in nature.
3. ALICE THE SMATER CHATBOT
Artificial Linguistic Internet Computer Entity (ALICE) was first ever implemented by
Wallace in 1995. Alice carries out the interaction with humans through heuristically pattern
matching rules to input provided by humans. The development began in 1995 by Richard
Wallace and was rewritten in Java in the early 1998. An XML schema AIML (Artificial
mark-up language) was used to specify the heuristic conversation rules. AILML also
comprises of AIML objects which consist of units called topics and categories which forms
the basic unit of knowledge in AIML. Each category comprises of a rule and set of patterns
that matches against the input of the user which generates the answer given by the chat bot.
ALICE consists of three AIML categories:
 Atomic Categories: Consist of patterns without the wildcard entries.
 Default Categories: Patterns that have wildcard entries are included in this category.
The symbols will although match some input but they will differ in their alphabetical
order.
 Recursive Categories: Comprises of templates which refer to recursive reduction
rules. The symbolic representation reduces the complex grammatical forms to the
simple forms, applies divide and conquer for the splitting of input to 2 or more
subparts.
A normalization process is applied to remove the punctuations before the start of the
matching process. AIML interpreter matches word by word to obtain the best longest
pattern. As soon as the match is found the process stops and the template belonging to the
processor is thereby processed by the interpreter to construct the desired the output.

4. RISE AND EVOLUTION OF CHATBOTS

There is no doubt in this that chat bots have become one of the fast automation tools of the
moment thereby driving the human computer interaction to a next level. Many individuals
assume that it’s a result of the AI hoopla created by Face Book to gap up its traveller
platform in order to build bots by the developers. Hence it can be said that chat bots have
become a sensation in a very short span of time but the upheaval in chat bots is a result of
several factors that came into existence from early 2000s to today.

4.1 GROWTH IN THE USAGE OF INTERNET

In the past few years consolidated itself as a powerful platform that brought out a modified
approach for the business we do and the way we communicate. The recent evolution in the
internet can be marked on the basis of 2 factors namely mobile technology and the social
web. Since its creation in year 2004, Facebook has big into a worldwide network of
over a pair of 230 million active users whereas mobile technology has created a potential
far larger reach of the web thereby increasing the amount of net users everywhere. The
success of a chat bot depends upon the factor that how many users will be accessing it
and internet provides handful numbers of platforms that will give a boost to the usage of
chat bots.
 4.2 RECENT ADVANCEMENT IN TECHNOLOGY
The information generated by web users will be if there aren't any tools obtainable to
leverage the information for learning functions. Within the past few years there has been
a boon for the sector of machine learning and AI. In the early years of 2000s, the
machine learning field evolved with addition of deep learning, which helps laptop
machines to “see” and perceive things in text, images, audio, or videos. The highest
technology firms pushed the event of AI to leverage.
The bot marked its evolution by shaping the internet as the first known search engine indexed
sites by helping the people in directing towards the information they are looking for. The
Information Technology industry has transformed in the recent years and so the user behavior
thereby making the chat bots more and more fascinating towards the software industries. The
advancement in chat bot messaging applications has exposed the ubiquity of connectivity
across the web.

5. COMPONENTS OF A CHAT BOT

As already known, Chat bot came into existence as a software program that chats naturally
and get the work done for humans. A chat bot comprises of different components:

5.1 Natural Language Processing (NLP): NLP is a component of artificial intelligence that
deals with the interaction between computers and human in order to process the large
amount of data in natural language. It can be also defined as automatic manipulation of
natural language. NLP application poses a challenge to develop as they require humans to
speak to them using a programming language which must be highly structured and precise
in nature. The natural language processing is carried out using two techniques namely
syntax analysis and semantic analysis. Syntax analysis is a process of analyzing the strings
of symbol in the natural language conforming to the rules of formal grammar. Semantic
analysis provides meaning of the words

5.2 Dialog Manager: The dialog manager decides what message will be conveyed to the user,
given its input and the interactions made by the user in the past

5.3 Content: It forms the third component and will describe what the bot is going to say once it
is decided what to say. The manner in which the content is structured will describe how the
content is going to be perceived by the user or client.

6. ARCHITECTURAL MODEL OF CHATBOT

The design model of a Chatbot is set to support the core purpose of development. There are
2 ways to attain a response from a chat bot, either by generating response from start as per
machine learning models or by using a heuristic approach to select the most promising
response from the library of already defined responses [7] .
6.1 GENERATIVE MODEL
This model is used to develop smart bots that are advanced in nature, but the model is rarely
used as it requires the implementation of complex algorithms.
6.2 RETRIEVAL BASED MODEL
This model is easy to build and is more reliable in nature. Here we know the possible type
of response that will be generated and can ensure that no incorrect response will be
generated by the chat bot.

The figure below listed as 6.1 and 6.2 describes the flow of chat in generative model and
retrieval based model.

User Message Previous Messages Content User Message

Response

Generative
Model
Retrieve Based
Model

Response

Fig 6.1 Generative Model Response

Fig.6.2 Retrieval based model

7. GENERATION MECHANISM OF RESPONSE BY CHAT BOTS

We will be discussing two different ways in which response are generated by the chat bots
to understand the human intent.

7.1 Artificial Intelligence Modelling Language (AIML)

It is an XML mark up language used to create schemas for artificial intelligence based
applications, developed by Alicebot free software community in 1995-2000. The
vocabulary of AIML has two special wildcard characters ‘*’ and ‘_’. The interpreter gives
more importance to patterns having ‘_’ rather than patterns having’*’. Patterns in AIML are
case insensitive. To run the AIML, setting up a Java development Kit is a primary
requirement. Categories, template and patterns forms the basic components of AIML.

7.2 Pattern Based Heuristics

Chat bots can generate response either by using classifiers of machine learning or
conditional statement if else. Artificial Intelligence Mark- up Language is used in the
 development of Chatbot for writing patterns and response. These bots dissect user message,
notice synonyms and ideas, tag components of speech and conclude that rule matches the
user question. But, these bots don't run machine learning algorithms or the other
Application programming Interfaces (APIs) until they are programmed specially.

7.3 Intent Classification based on Machine learning

Here the bots are trained using machine learning. The chat bot is trained to pick up a pattern
of data from a training set of thousand of already existing patterns that are likely to be faced
by chat bot.
The response generated depend on user, it can be a simply generated response or could be
in some template based on users intent put in some with variables.

8. TYPES OF CHAT BOTS

In this era of application, mobile applications have been take over by chat bots through their
ability to imitate conversations and provide instant digital connections.

Fig.4 Types of Chat bots

Chat bots once installed are capable of handling queries at any time. Chat bots are helping
the organization to keep pace up with the trend and helped the organizations to acquire
customer satisfaction.

The below presented Table 1 will depict a short summary of all the chat bots that have been
listed in the figure which are scripted based chat bots, NLP based chat bots, social
messaging chat bots. Service – Action chat bots and voice enabled bots.
 Table 1: Description of the chat bots

Type of Chat Bot Description

Scripted Chat Bot In the scripted chat bot, the interaction with the
client takes place using a predefined set of
queries. Either a questionnaire about what the
customer is required to ask is used to design
the chat bot or the rule based chat bot is created
where each action performed by the user apt
the bot to respond.

NLP Chat Bot Artificial intelligence based chat bots uses

natural language processing as a prominent
technology at its core. NLP is used to map
user’s text or voice to intent. It also performs
the classification of messages and then parsing
is performed in order to achieve variables for
answer
Service- Action Chat Bot From the user, relevant data is collected in
order to perform or complete the action.
Airline Business can be an example of such
type of bot
Social – Messaging chat bots These bots have their integration within the
social messaging platform, thereby making it
easy for the users to directly interact with the
bot

9. WORKING MECHANISM OF CHATBOTS

Chat bots are somewhat similar to the applications we use in daily scenario as they contain a
database, an application layer and a number of application package interfaces (APIs) to
connect to an external call [10]. The intention of the user cannot be directly understand by
the bot hence they are first trained with the data.

The customer support chat bots are installed with several conversation logs which helps them
in understanding what type of response need to be generated to a particular user query. Smart
feedback loops can be implemented once the Chatbot is live and started interacting with the
people. Three classification methods are adopted by the chat bots.

9.1 PATTERN MATCHERS

Pattern matching is used for the classification of text in order to generate a suitable
response. The standard structure for these patterns is Artificial Intelligence Markup
Language. The chat bots will only respond according to the associated pattern and can’t
go beyond that.
 Example:
<aiml version=”1.0.1” encoding=”UTF-8”?>
<category>
<pattern> WHO IS NELSON MANDELA</pattern>
<template> Nelson Mandela was the President of South Africa from 1994 –
1999</template>
</category>
</aiml>

9.2 ALGORITHMS
A unique pattern must be available in the database for the response generation to each
question. Algorithms are used to generate a manageable structure by reducing the
number of classifiers. With a new input sentence, each word is counted for how many
times it has occurred and a score is assigned to the class. The class with the highest score
is likely to be related to the input sentence.

Example of a training set

Class: greeting
“Good Morning”
“How are you?”
“Good Afternoon”
“Hello”

Input Sentence classification

Input: “Hi Good Morning”
Term: “Hi” (no matches)
Term: “Good” (class: greeting)
Term: “Morning” (class: greeting)
Classification: greeting (score=2)

10. NATURAL LANGUAGE PROCESSING(NLP) FOR CHATBOT

NLP deals with the interactions between computer systems and humans in order to process
large amount of information. Chat bots such as Amazon’s Alexa or Siri will be considered
inefficient without NLP, as it forms the basic unit that allows the chat bot to understand,
interpret the user message so that an appropriate response could be generated. Whenever a
message like “Good Morning” is sent it is the NLP that let the chat bot know that a standard
greeting have been posted and thereby an appropriate response is generated by the chat bot.

Machine Learning is one of the best approaches of NLP that can be used to train the bots.
The NLP based bots generates less false positive outcomes, identify user input failures and
uses comprehensive communication process to generate user responses.

According to Maruti Techlabs, there are certain capabilities that a Natural Language
Processing must have.
  CAPITALIZATION: Removal of capitalization from common nouns and recognition
of proper nouns.
 EXPANSION OF VOCABULARY: Providing addition of synonyms and expansion of
chat bot vocabulary with the help of machine learning.
 CONTRACTIONS: Simplification of task processing and removal of contractions.
 DIGITS v/s NUMERIC WORDS: Recognition of communication of numeric values as
words or digits
 VOCABULARY TRANSFER: Transferring the developed vocabulary from one chat
bot to another chat bot
 SINGULAR v/s PLURAL NOUN: Processing of singular and plural nouns should be
done in a similar way.
 TENSED VERBS: Communication of a single verb in different tenses should be
treated as synonymous.
 MESSAGE PERSONALIZATION: Replacement of default universal responses with
unique configured messages.

11. TRENDING ARTIFICIAL INTELLIGENCE PLATFORMS

Evolution of chat bots have revolutionised the customer experience. Therefore chat bots are
considered as the technology of new generation. There are several platforms available for the
development of chat bots.

We have created accounts on few chat bot platforms to give a better visualization of how the
appearances of the platforms look like.

CHATFUEL: It provides the creation of smart and intelligence AI based chat bots. It provides
answers automatically to the frequently asked questions from your customers. This automation
ensures to provide constant assistance to the users. Chat bot is a leading chat bot platform for
face book messenger and around 46% of the messenger bot are running on the chat fuel

LANDBOT.io: This tool allows creating personalised conversational chat bots in order to
provide interaction with the prospective customers. It provides the bots to get published in
different formats and it can be launched on multiple channels. Land bot also generates
memorable experiences from the static lead generation. It allows engage of potential customers
with a help of a user friendly editor. One can easily customize the bot with their brand entity.
Figure 5 describes the appearance of the platform.
 Fig.5 First appearance of Landbot.io

MEYA.AI: It comprises of a live debugger, code editor and a visualize tool. It provides scaling
of bots along with along with training and hosting. Meya also provides integration with third
party applications to provide easy usage.
XENIOO: Xenioo does not require any line coding .It allows instant creation and publish of
intelligent chat bots for the social media platforms. The chat bots are trained using intent and
expressions. Figure 6 depicts how the first look at Xenioo platforms looks like. The very next
figure 7 describe about the assistance Chatbot Lisa in Xenioo.

Fig .6 First look of the Xenioo Dashboard after the sign in process
 Fig.7: Lisa the assistance chat bot in Xenioo

BOTSIFY: It allows us to create for websites or the facebook messenger application and also
provides creation of automated bots online. It provides faster support to the business and creates
multi step question answer sequence. It creates context aware stories.Figure 8 describes the bot
assistance offered by the Chatbot in the Botsify platform.

Fig.8: Bot assistance by Botsify that helps you to get started.

 TARS: Tars help to observe a page experience in a more personalised manner by letting
you know who clicks on advertisement which results in the increase in conversion rate. This
platform provides several lead generation templates and for several industries which
includes healthcare, insurance, legal and more.

SEQUEL: It allows creation of customize chat bot templates using its drag and drop
embedded editor. It designs chat bots that interacts with users on a conversational level
using natural language processing technology.

PANDORABOTS: Pandora bots is a huge platform for building bots but requires initial
coding skills. It offers flexibility and a solid interface that speed up the development
process.

MOBILEMONKEY: It allows creation of bots for face book messenger that drives the
marketing on face book onto a next level. It can be used to manager large amount of
business lead activities.

12. CONVERSATIONAL USER INTERFACES

Conversational Interfaces allows the interaction on human terms with the computer system. It
is an interface that uses plain language to talk and write. Interface is considered as a great
part of user experience and thereby conversational interfaces form the latest trend of digital
designing. Conversational UI are available in 2 forms a Chatbot that allows you to type and a
voice assistant for speech [19] .Voice assistants launched by Amazon, Google and other
technical giants are becoming smarter day by day. The conversation User interfaces are
classified into three categories
 Voice based assistants
 Text based assistants
 Basic Bots

12.1 BASIC BOTS

They form the main building blocks for the interface and here inputs given are rather
restricted in nature. Deployment of basic commands and basic inputs are allowed for greater
chat bot integration.

12.2 TEXT BASED ASSISTANTS

In text based chat bots texting is regarded as the primary mode of communication which
could involve media and other elements of user interface. The replies produced by the chat
bot will depend upon the quality of the input given. Text based chat bots have extensive
library but since the vocabulary depends upon the industry so the accuracy of the Chatbot
may be hindered.

12.3 VOICE BASED ASSISTANTS

Voice assistant are majorly adopted by the shopping websites. Voice based assistants
provides greater accuracy because of the greater integration. They are providing greater
automation hence reducing the time in obtaining a seamless experience.
13. BRICKS OF BOT BUILDING

Today the chat bots not only understands the commands but also the human spoken
language which has been possible due to the natural language processing feature exhibited
by the artificial intelligence which has made the chat bots smarter with time. The table
below provides an overview of intents, entities that lays the basic foundation of bots
building.
Table 2: Bricks of Building Bots

BRICKS OF BUILDING BOTS

ENTITIES They are described as keywords or phrases that
are searched by the chat bot in user message.
The entities helps the chat bot in identification
of the subject matter of the conversation, thus
providing a great experience

INTENTS Intent is identification of user’s intention who

is using the chat bot. Text classification helps
in detecting the intent of the user message and
also classifies the sentences into multiple
classes.

14. DESIGN PRINCIPLES OF CHATBOT

Chat bots differ from humans as they are designed with an aim of assisting the user with what
they want. They have transformed the consumer technology by providing a human like
experience. There are certain design principles that one should keep in mind while designing the
chat bots.

KEEP IT SIMPLE AND SHORT

The conversation of chat bots should follow the linear conversation flow and it must be bounded
to particular subjects. Even though if we have a lot to say, we should it keep it as precise as
possible. Ignore the detailed presentation of the message; we can break the message into
fragments.

THE CHAT MEDIUM

The source channel of the message is known whenever the message is sent to the Chatbot. The
bot interaction is all about call and response and we should keep in mind that bot will not be able
to modify the conversation the way humans could do it. Each platform provides several
messaging platforms and it depends upon the user what formatting they want to use.
OPTIMIZATION FOR THE END USERS
Bots must be designed to improve the user end experience. The UX developers must focus to
improve the areas where humans are not good, not on the areas where humans are already doing
well.

15. DESIGNING CHAT AND VOICE BOTS

In today’s digital era, Alexa and Siri have become a part of about 20% of households thereafter
making a big way to enterprise and daily work life. Creating the right user experience is
important during the design phase of the bot. Voice based chat bots create an expression of how
the interaction is governed between the bot and the human. Thus the design of the Chatbot is all
about creating the right experience for the user. The below figure describes what are the steps
that are followed by the organization Maven wave who are expertise in experience design.

Fig. 9 Design Process Flow

 A conversational process must be followed while designing the chat bots. Follow a
methodology that result in delivering good experience design.
 The designing process of the bot should be in the same way as we train a new employee.
In the situation of a real world machine learning is being used extensively in providing
training to the bots.
 The platform being selected for the designing process plays a crucial role in the designing
process as they control the quality of the conversation along with the ability to extend the
functionality.
 Along with a good conversational design a right access to data and system is responsible
for the success of the chat bots.
16. BENEFITS OF CHAT BOTS
Chat bots are rising at a pace and are being adopted worldwide. They have launched new
ways of interaction with the rise in artificial intelligence technologies. Following are the
benefits chat bots have successfully provided to the users of each industry and domain [24].

Fig 10 List of benefits provided by the chat bots

 There is an increase in the number of users of messaging applications as

compared to the users of the social media. Chat bots have provided real time
assistance similar to a real store.
 Chat bots are providing 24*7 assistance to the users hence offering great customer
satisfaction.
 Chat bots have revolutionised the marketing strategies completely and forms the
customer service strategy.
 Good customer relationships are built as chat bot keep a track of the interest of the
customer and helps them in providing suitable recommendations.
 Chat bots require consistent evaluation and optimization where optimization
includes setting of appropriate goals and broadcasting.

17. CHATBOTS: OFFERING A BOOM TO BUSINESS

The Google CEO, Mr. Sundar Pichai told “At a very long run, we would be evolving from a
mobile-first to and AI- first world”. Chat bots and virtual assistants which came into existence
as a niche concept has now become the need of the hour for every enterprise. With an improve
in sales and customer service chat bots have transformed the business. It acts as a bridge
between the brand and the customer by providing 24 hours constant assistance. They establish
personalised customer interactions by catching the attention of the customer through past
interactions. Organizations are able to handle more tasks and also scaling up their operations on
 a global market because of the chat bots. In the era of a competitive business no organization
can imagine a passive communication; chat bot are increasing the customer engagement and
reduces the cost of carrying out the customer support operations.

WHY THERE IS A NEED TO ANALYSE

Building a great analytics is an important step required to construct a nascent technology. As
long as we don’t know the usage patterns it is difficult to predict what you user like or dislike.
Analytics is the science that deals with the extraction of patterns and trends from the data that is
available from varied sources. The data today in real world exist in the form of structured,
unstructured and semi-structured data. Today messaging applications like Face book and
Whatsapp have information about who are our friends and the members of our family thereby
providing a way to be available online and connect with the people worldwide. The data tells the
usage of the application and when interpreted and worked upon can be a success indicator.
Certain business metrics must be kept in mind to mark the progress of an effective business. A
business metric is a quantifiable measure that tracks the status of the business. The toolkit that
measures the growth of chat bots is at a nascent stage. The criterion that measures the growth of
messaging applications differs from the criteria that are used to measure the progress in chat
bots. Chat bots works well in the use cases of customer support and but if n case a drop off
occurs one need to monitor the message where the drop has occurred and must mined the pattern.
While measuring the analytics absolute number is not concerned the concern point lies in the
relative change in number.

18. PROGRAMMING LANGUAGES

Chat bots are becoming unavoidable part of human life. The choice of the programming
language will depend upon the platform that is being used to develop the chat bot. Below we
present some of the common programming languages that can be used in the development
process.

PYTHON
Python is a high level programming language created by Guido van Rossum and was released in
1991. It relies on indentation and uses new lines for the completion of a command, used in
scripting and web designing.

CLOJURE
It is a functional programming language that runs on Java Virtual Machine and is a dialect of the
programming language LISP. Instead of side –effect based looping it highlights the recursion and
high order functions.

JAVA
It provides all the high level features that are required by the artificial intelligence projects. It
runs on java virtual machine and provides a sophisticated framework using visualizations.
19. DIALOG FLOW CHATBOT FRAMEWORK
Dialog flow provides new ways to interact with the users with the help of voice and text based
conversational user interfaces. It incorporates Google’s machine learning technology and runs on
Google cloud platform. We have presented a pictorial representation of the process of intent and
entity creation in the dialog frame work by creating a Google account. Also the agent created by
us is named as AMARA

 Sign up on the dialogflow.com using a Google account.

Fig. 11 Sign to dialog flow using Gmail account

 Accept all the requested permissions and then you will be allowed to access the console.
 Create a new agent, the language primarily would be English, but you can modify it
according to you. Here the new agent created is AMARA depicted by figure 12.

Fig.12: Amara the agent created

 First intent created as shown in figure 13

Fig.13: Hello is the first intent created

 Create Entities. In figure 14 created entity is WORK.

Fig.13 Entity created work

These steps are a basic introduction to the framework Dialog flow, how you can use it.

20. BUILDING A CHAT BOT WITH PYTHON

The figure below will first describe what are the basic steps required to build the bots.

Fig 14: Steps required building the bots

 Creating a new project. We have created the project as AMANDA CHATBOT as shown
in figure 15. Here IDE is Pycharm

Fig.15 Creation of the new project

 After creating the project, create a new python file. Amanda Chatbot> right click> new>
python file shown by figure 16

Fig. 16 Create a new python file

 A file named conversation.py has been created as depicted by figure 17.

 Fig.17 Conversation.py is the python file

 Creating the text file for AMANDA CHATBOT. Amanda Chatbot> new>file. Figure 18

Fig.18 Creation of text file

 This is how our text file appeared as shown by figure 19.

Fig.19 Content of the text file

 Installing the package for the AMANDA CHATBOT

 Fig.20 List of available packages

Importing the modules of chatterbot. It is a python library that generates automated response for
the users input.Following commands will be used

From chatterbot import Chatbot

From chatterbot.trainers import List Trainer
Bot=Chatbot(“ Chatbot”)
Open the conversation file with the command conversation
open('Textfile.txt','r').readlines()

 Finally train the bot and put the conversation together.

21. CHATBOT IN FINANCE

Artificial Intelligence has completely changed the outlook of the people looking at the things
around. It now depends upon the organization how they can explore the capabilities of the chat
bot to explore their business on the front end. Chat bots are preferred by the customer themselves
as there is no longer a need to be in a queue at the helpline, one can easily get in touch through
the messaging.
Financial sector have transformed completely with the entrance of artificial intelligence as chat
bots have become a major technology transforming the customer service automation. Below are
the advantages that chat bots have provided in the financial industry.
Reduction in call centre agents
Increase in margins and cost savings.
Easy to handle
Personalization of content
Scalable

FinChat bot is an AI powered chat bots for the financial industry. Holly is the virtual assistant
that helps in the customer interaction. Holly can interact with several potential clients in different
languages anytime. It can be deployed across multiple platforms. Below we present a short
interaction with the chat bot Holly in the figure 21

Fig 21 Holly the bot

22. CHAT BOT IN HEALTHCARE

Healthcare is another field where chat bots and health are working great together. Chat bots for
healthcare can either be a counsellor or a care giver depending upon the functionality of the chat
bot. They are helping in providing augmentation and diagnosis. As the healthcare services are
becoming patient-centric, providing personalised and satisfactory experience has become a
priority for the health care providers.

As the medical knowledge keeps on updating so the need of proper understanding of the
technology is required. The chat bot SafedrugBot provide with the right information about the
drug dosage.
Florence another bot launched in 2017 helps in reminding the patients to take pills, track their
weight, periods and moods. It also provides the location of nearby pharmacy or a doctor
depending upon your disease. Below is the representation of the chat with Florence chat bot by
figure 22

Fig. 22 Florence – Medical assistant bot

23. FUTURE OUTLOOK WITH CHAT BOTS

Chat bots have made a great impact on each and every industry and operational domain.
Customer service is the major area hit by the chat bot. With the introduction of devices such
as Google Home and Amazon Echo, artificial intelligence has already hit a major area of our
daily lives, thereby growing the space of artificial intelligence in workspace [26]. Chat bots
are providing a seamless engagement experience to customers across all the platforms. They
have enhanced interaction and also helped in managing large amount of information.

Continuous improvement in the natural language processing has enabled the bots to have a
conversation like the human being. Performing simple jobs in a repetitive and efficient
manner have made the chat bots to build an organization fatigue-free. They have made the
human agents to handle complex queries. The global Chatbot market is expected to increase
by $1.34 in valuation by 2024. Chat bots have come up with a maximum number of
opportunities that offers personalization. For ensuring 100% of customer satisfaction, we can
not rely in technologies completely at some point human intervention is required, But still
chat bots provide a logical, transparent and clear communication. So if we take into account
the current market trends then we can say that chat bot have a great future ahead.
REFERENCES
1. Shawar, B. A., & Atwell, E. (2007, January). Chatbots: are they really useful?. In Ldv forum
(Vol. 22, No. 1, pp. 29-49).
2. Dale, R. (2016). The return of the chatbots. Natural Language Engineering, 22(5), 811-817.
3. Khan, R., & Das, A. (2018). Introduction to chatbots. In Build Better Chatbots (pp. 1-11).
Apress, Berkeley, CA.
4. Klopfenstein, L. C., Delpriori, S., Malatini, S., & Bogliolo, A. (2017, June). The rise of bots:
A survey of conversational interfaces, patterns, and paradigms. In Proceedings of the 2017
Conference on Designing Interactive Systems (pp. 555-565). ACM.
5. Ferrara, E., Varol, O., Davis, C., Menczer, F., & Flammini, A. (2016). The rise of social bots.
Communications of the ACM, 59(7), 96-104
6. Abdul-Kader, S. A., & Woods, J. C. (2015). Survey on chatbot design techniques in speech
conversation systems. International Journal of Advanced Computer Science and Applications, 6(7).
7. Cahn, J. (2017). CHATBOT: Architecture, design, & development. University of Pennsylvania
School of Engineering and Applied Science Department of Computer and Information Science.
8. Cahn, J. (2017). CHATBOT: Architecture, design, & development. University of
Pennsylvania School of Engineering and Applied Science Department of Computer and
Information Science.
9. Bird, S., Klein, E., & Loper, E. (2009). Natural language processing with Python: analyzing
text with the natural language toolkit. " O'Reilly Media, Inc.".
10. AbuShawar, B., & Atwell, E. (2015). ALICE chatbot: Trials and outputs. Computación y
Sistemas, 19(4), 625-632.
11. Kerlyl, A., Hall, P., & Bull, S. (2006, December). Bringing chatbots into education: Towards natural
language negotiation of open learner models. In International Conference on Innovative Techniques
and Applications of Artificial Intelligence (pp. 179-192). Springer, London.
12. Rahman, A. M., Al Mamun, A., & Islam, A. (2017, December). Programming challenges of
chatbot: Current and future prospective. In 2017 IEEE Region 10 Humanitarian Technology
Conference (R10-HTC) (pp. 75-78). IEEE.
13. Følstad, A., & Brandtzæg, P. B. (2017). Chatbots and the new world of HCI. interactions, 24(4), 38-
42.
14. Burden, D. J. (2008, December). Deploying embodied AI into virtual worlds. In International
Conference on Innovative Techniques and Applications of Artificial Intelligence (pp. 103-115).
Springer, London.
15. Shum, H. Y., He, X. D., & Li, D. (2018). From Eliza to XiaoIce: challenges and
opportunities with social chatbots. Frontiers of Information Technology & Electronic
Engineering, 19(1), 10-26.
16. McTear, M. F. (2002). Spoken dialogue technology: enabling the conversational user
interface. ACM Computing Surveys (CSUR), 34(1), 90-169.
17. Pearl, C. (2016). Designing Voice User Interfaces: Principles of Conversational Experiences.
"O'Reilly Media, Inc.".
 18. Radziwill, N. M., & Benton, M. C. (2017). Evaluating quality of chatbots and intelligent
conversational agents. arXiv preprint arXiv:1704.04579.
19. Hill, J., Ford, W. R., & Farreras, I. G. (2015). Real conversations with artificial intelligence:
A comparison between human–human online conversations and human–chatbot
conversations. Computers in Human Behavior, 49, 245-250.
20. Janarthanam, S. (2017). Hands-on chatbots and conversational UI development: Build
chatbots and voice user interfaces with Chatfuel, Dialogflow, Microsoft Bot Framework,
Twilio, and Alexa Skills. Packt Publishing Ltd.
21. Paikari, E., & van der Hoek, A. (2018, May). A framework for understanding chatbots and
their future. In Proceedings of the 11th International Workshop on Cooperative and Human
Aspects of Software Engineering (pp. 13-16). ACM.
22. Zumstein, D., & Hundertmark, S. (2017). CHATBOTS--AN INTERACTIVE
TECHNOLOGY FOR PERSONALIZED COMMUNICATION, TRANSACTIONS AND
SERVICES. IADIS International Journal on WWW/Internet, 15(1).
23. Chung, K., & Park, R. C. (2018). Chatbot-based heathcare service with a knowledge base for
cloud computing. Cluster Computing, 1-13
24. Brandtzaeg, P. B., & Følstad, A. (2017, November). Why people use chatbots. In
International Conference on Internet Science (pp. 377-392). Springer, Cham.
25. Zamora, J. (2017, March). Rise of the chatbots: Finding a place for artificial intelligence in
India and US. In Proceedings of the 22nd International Conference on Intelligent User
Interfaces Companion (pp. 109-112). ACM.
26. Khan, R., & Das, A. (2017). Build better chatbots: a complete guide to getting started with
chatbots.

View publication stats