See discussions, stats, and author profiles for this publication at: [Link]
net/publication/371491011
CHATBOT ADOPTION IN TRAVEL AND TOURISM SERVICES
Article · June 2023
CITATIONS READS
5 3,398
3 authors, including:
Kirti Muley
Bahrati Vidyapeeth's Institute of Management and Information Technology
33 PUBLICATIONS 20 CITATIONS
SEE PROFILE
All content following this page was uploaded by Kirti Muley on 12 June 2023.
The user has requested enhancement of the downloaded file.
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
CHATBOT ADOPTION IN TRAVEL AND
TOURISM SERVICES
Ms Khan Zainab Irfan
Prof [Link] Muley
1Student 2Professor
Master’s in Computer Application (MCA),
Bharati Vidyapeeth’s Institute of Management and Information Technology, Navi Mumbai, India
________________________________________________________________________________________________________
Abstract: Chatbots have gained increasing importance for research and practice with a lot of applications available
today including Amazon’s Alexa or Apple’s Siri. In this paper, we present the underlying methods and technologies
behind a Chatbot for e-tourism that allows people textually communicate with the purpose of booking hotels, planning
trips, and asking for interesting sights worth being visit. In particular, we show how model-based reasoning can be used
for enhancing user experience during a chat, e.g., in cases where too many possible selections are available or where
user preferences are too restricted causing inconsistencies and as a consequence not possible answers to be provided.
Besides the underlying foundations, we provide a use case from the intended tourism domain to show how such a model-
based chatbot effectively can be used in practice.
Index Terms - Machine Learning, Chatbot, Natural Language Processing, Pytorch , Tokenization, keywords
matching, Sentiment Analysis, Parsing, Chatbot adoption in travel and tourism services.
_____________________________________________________________________________________________________
I. INTRODUCTION
A survey has shown that 87 % of users would interact with a travel chatbot if it could save them time
and money. Chatbots act as personal travel assistants to help customers browse flights and hotels,
provide budget-based options for travel, and introduce packages and campaigns according to
consumers’ travel behaviour.
Chatbots are a form of artificial intelligence system that allows a human-computer interaction in a
natural language form. They could be based on rule sets or neural networks in order to decide the correct
answer to the user’s request. Chatbots are not restricted to certain application domains. They are flexible
enough to be used in many different application scenarios and domains including systems for tourists
recommending sights, hotels, or even complete travel plans. Often chatbots rely on pre-specified
patterns that trigger the chatbot’s behaviour, restricting its space of interaction with users. Travel
chatbots, often called “virtual travel agents,” are now emerging as a tool which can learn about its
consumer and take on action items, such as booking, or searching through booking emails and calendars
to build an itinerary. Some areas of the tourism industry, particularly airlines, are leading the way in
the adoption of chatbots such as Alex (United Airlines), Mildred (Lufthansa) and Finn (Finnair), but
many others have been very slow to adopt this strategy.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 25
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
II. LITERATURE REVIEW
1. Literature review: 01
Title Author Working Methodology
Chatbot-based Iulia Nica, The chatbot could 1. EntRecom takes a
Tourist Oliver A. Tazl randomly offer knowledge base, a set of
Recommendations and Franz alternative features to customer requirements, and the
Using Modelbased Wotawa. choose from, or even maximum number of
Reasoning better, it could recommendations,
and computes all
optimize the searching
recommendations.
process by making use
of an algorithm to 2. MI REQ starts with
determine the attribute, assuming one faulty requirement
which if constrained, (i = 1) and continues to search, if
leads to the largest necessary, up to the number of
information gain. existing requirements. The
constraint solver is this time
called restricting the solutions to
the specific cardinality.
3. GetBestEntrAttr iterates
over the set of attributes. In every
step it calculates the entropy for
the current attribute.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 26
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
2. Literature review: 02
Title Author Working Methodology
Designing a Prof. Ioannis Have provided the A chatbot, created for the
Chatbot for Magnisalis background regarding Messenger app, that can be
the tourism industry executed. However, its
Tourism
and how potential capabilities at the moment are
travellers are searching limited and some messages do
when planning their not display the content we want.
next trip, either for
Let’s have a more detailed look
business or for
vacation. We have also at every step of the chatbot. The
seen some facts about answer of the user can be
chatbots and artificial completely different from what
intelligence, some we have provided. This will be
implementations of solved with the use of Natural
chatbots from Language Processing (NLP) and
companies in the with the appropriate training of
tourism industry, and the chatbot. We can train the
the benefits in general, chatbot by providing it with
that bots bring in this different possible examples of
industry.
replies that it will encounter.
III. OBJECTIVES
This paper discusses Artificial Intelligence (AI) but is mostly focused on chatbots. The paper
introduces the fundamentals of chatbots and their functionality and conversational interfaces. Finally,
the thesis focuses on providing information to evaluate digital trends and growing demand in the
tourism industry. The primary objective of this research is to design and create a chatbot for
holidaymakers to plan their trip and help them to make the right choices. More specifically, the
implemented chatbot aims to help travellers to organize their itinerary and give them personalized
suggestions based on their preference. To solve the research goals, the study has the following
questions:
1. How does a travel chatbot provide good customer satisfaction?
2. How do users feel about chatbot technology, and how can chatbots improve tourists’
engagement?
Solving these questions helps in designing and creating an advanced chatbot that offers a wide range
of options to travellers, which leads to more customer satisfaction. Many travel companies, airlines,
and hotels may use this research in order to attract more clients and improve their customer
experiences.
A chatbot help potential travellers plan their next trip in a few easy steps. This chatbot can help the user to
suggest where to visit and show where are recommend places. User can also use this chatbot to find hotels
and type of transportation around user's location.
It will help the user selects his/her destination:
• Find and book a transportation method to the destination.
• Find accommodation.
• Find things to do and visit there.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 27
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
IV. SCOPE
The reason I chose this topic was due to the recent advancement in chatbot technology and machine learning
and its impact on the tourism industry. The scope paper focuses on chatbot’s ability to provide options and
plans for travellers and obtaining more tourists for businesses. Finally, rather than diving into the technical
specifics of machine learning like technologies or statistical 3 models, the paper discusses the origin and
history of machine learning, artificial intelligence, and chatbots. Also, it gives limited details regarding the
implementation of a chatbot by programming. Deployment of chatbots has the following advantages: bots
assist users in navigating websites, thus shortening time and assisting in quick decision-making; build social
relationships with customers; maintain customer confidence in firms; and strengthen customers’ emotional
bonds with firms. Additionally, scholarly evidence supports customers’ acceptance of robots in hotel services
in particular with chatbots significantly increasing hotel sales
Some areas of the tourism industry, particularly airlines, are leading the way in the adoption of chatbots such
as Alex, Mildred and Finn, but many others have been very slow to adopt this strategy.
V. PROCESS DESCRIPTION / METHODOLOGY ADOPTED
• Natural Language Processing:
NLP is of particular importance for chatbots because this technique determines how the bot
will understand and interpret the text input. The goal of an ideal chatbot would be to converse
with the user in such a way that the user is completely unaware that they are talking with a
machine. This algorithm attempts to learn through machine learning and an abundance of
conversational data, the intricacies of human language. NLP helps the bot understand text data,
comprehend grammar, sentiment and intent. This is primarily due to the wide range of
functionalities offered by NLP such as text summarizations, word vectorization, topic
modelling, PoS tagging, n-gram and sentiment polarity analysis. Due to the nature of NLP
techniques to try and mimic human conversation, testing this algorithm is essentially testing
the communication abilities of the chatbot.
Fig.1 NLP
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 28
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
Testing NLP algorithm is very subjective and there is no absolute standard or benchmark to assess
the algorithm’s performance. Some of the most popular and appropriate ways to test NLP
algorithms are user satisfaction and feedback analysis. Conducting data analysis and text analysis
on user satisfaction ratings may be simple but they do not provide enough insight into the results
and are therefore not recommended. Variations of Turing test can be used to test the algorithm as
NLP tends to try and mimic human linguistic conversations. The Turing test can be used to assess
how close the algorithm is to mimicking actual human interactions and human language. The test
could also involve users trying to differentiate between conversations with that of a human and to
that of a NLP based chatbot as shown in Fig.(1). Natural Language Processing includes the
following steps:
1. Tokenization: The NLP separates a series of words into tokens or pieces that are
linguistically representative, with a different value in the application.
2. Sentiment Analysis: It will study and learn the user’s experience, and transfer the inquiry
to a human when necessary.
3. Normalization: This program model processes the text to find out the typographical errors
and common spelling mistakes that might alter the intended meaning of the user’s request.
4. Named Entity Recognition: The program model of chatbot looks for different categories
of words, similar to the name of the particular product, the user’s address or name,
whichever information is required.
5. Dependency Parsing: The Chatbot searches for the subjects, verbs, objects, common
phrases and nouns in the user’s text to discover related phrases that what users want to
convey.
• Pytorch:
PyTorch is a Python-based scientific computing package that uses the power of graphics
processing units (GPU). Since its release in January 2016, many researchers have continued to
increasingly adopt PyTorch. It has quickly become a go-to library because of its ease in building
extremely complex neural networks. It is giving a tough competition to TensorFlow especially
when used for research work.
Some of the key highlights of PyTorch includes:
1. Simple Interface: It offers easy to use API.
2. Pythonic in nature: This library, being Pythonic, smoothly integrates with the Python data
science stack.
3. Tensors: It is basically the same as a NumPy array. To run operations on the GPU, just cast
the Tensor to a Cuda datatype.
4. Computational graphs: PyTorch provides an excellent platform that offers dynamic
computational graphs.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 29
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
VI. PROCESS DESIGN
Fig.2. Process design
There are seven steps to design the Chatbot process. The Chatbot design process figure (2) is shown
above: The first step to designing the Chatbot is to know the scope and requirements like why chatbot,
platform to launch chatbots and its limitations. The second step is to identify the inputs from users in the
form of queries through text, voice or images, from devices, and intelligence systems. The third step is
to understand the User Interface (UI) elements, that we can see in our applications.
UI elements are of five types they are: Command Line (CL), Graphical User Interface (GUI), Menu-
Driven Interface (MDI), Form-Based Interface (FBI) and Natural Language Interface (NLI).
After understanding user interface elements, the next step is to craft the first interaction and build a
conversation. The final step of the Chatbot design process is testing, which is done on mobile and
websites to know how it’s working.
VII. SYSTEM ARCHITECTURE
Fig.3.(a) System Architecture
In fig.3(a) the user inputs the query in the user interface in the form of text. The user interface receives
the user queries and then forwards it to the Agricultural Chatbot. In the Agricultural Chatbot, the textual
query undergoes a pre-processing stage. Preprocessing steps include Tokenization where the query
sentence is tokenized into words, then the stop words are removed, and then the words are stemmed to
their root words.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 30
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
• Response Retrieval using Machine Learning:
Fig.3 (b) Response Retrieval
As shown in fig.3 (b) retrieval-based chatbots work on the principle of graphs or directed flows. The chatbot is
trained to provide the best possible response from a database of predefined responses. Retrieval-based chatbots
use techniques like keywords matching, machine learning or deep learning to identify the most appropriate
response.
VIII. RESULTS AND DISCUSSION
• Results
Fig.4(a) Result
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 31
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
Fug.4(b) Result
Fig.4(c) Result
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 32
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
Fig.4(d) Result
The study aims to examine the development of a tourism Chatbot and focuses on the ability of the Chatbot which
is a text-based application for the tourism sector in the Active Beach zone of Thailand. The Chatbot is always
available to users, which can provide up to ate information about places, events weather conditions and provide
recommendations. We gave a technical overview of materials needed to build a Chatbot which are Natural
Language Processing (NLP), Machine Learning (ML), and ChatterBot. The paper also discusses how the
Chatbot classified, processed and made a prediction based on available data to find the best match by
using a machine-learning-based conversational dialogue engine build in Python.
IX. CONCLUSION
The chatbot is a tool that facilitates the provision of answers to (frequently asked) questions and stimulates user
action. It should not replace human-human interaction but support it. A chatbot should not be created to shield
the editor from communicating with the authors, but as a tool to provide faster response when the editor cannot
do this. In addition, the editor needs to be in control of the communication between the chatbot and the users.
ACKNOWLEDGMENT
I thank Prof. Dr. Kirti Muley, Dept. of Master in Computer Application in Information Technology for
her valuable support throughout the project, guiding us from time to time and looking into the project
when needed. We also thank python community for their user-friendly and extensive libraries.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 33
[Link] © 20XX IJCRT | Volume X, Issue X Month Year | ISSN: 2320-2882
REFERENCES
[1] Chrysovelidis, Georgios. "Designing a chatbot for tourism." (2020). (pp. 110-116).
[2] Nica, Iulia, Oliver A. Tazl, and Franz Wotawa. "Chatbot-based Tourist Recommendations Using
Model-based Reasoning." ConfWS. 2018. (pp. 334-339).
[3] Ukpabi, Dandison & Aslam, Bilal & Karjaluoto, Heikki. (2019). Chatbot Adoption in Tourism Services:
A Conceptual Exploration. 10.1108/978-1-78756-687-320191006.
[4] Vijayaraghavan, V., and Jack Brian Cooper. "Algorithm Inspection for Chatbot Performance
Evaluation." Procedia Computer Science 171 (2020): 2267-2274.
[5] Nica, I., Tazl, O. A., & Wotawa, F. (2018, September). Chatbot-based Tourist Recommendations Using
Model-based Reasoning. In ConfWS (pp. 25-30).
[6] Amalia, A. and Suprayogi, M., 2019, October. Engaging millennials on using chatbot messenger for
eco-tourism. In Third international conference on sustainable innovation (pp. 484-487).
[7] Ivanov, S.H., 2020. The first chatbot of a tourism/hospitality journal: Editor’s impressions. Hospitality
Journal: Editor’s Impressions (March 1, 2020). European Journal of Tourism Research, 24, p.2401.
[8] Kasinathan, V., Abd Wahab, M.H., Idrus, S.Z.S., Mustapha, A. and Yuen, K.Z., 2020, April. Aira
chatbot for travel: case study of AirAsia. In Journal of Physics: Conference Series (Vol. 1529, No. 2, p.
022101). IOP Publishing.
[9] Calvaresi, Davide, Ahmed Ibrahim, Jean-Paul Calbimonte, Roland Schegg, Emmanuel Fragniere, and
Michael Schumacher. "The evolution of chatbots in tourism: a systematic literature
review." Information and Communication Technologies in Tourism 2021 (2021): 3-16.
[10] Alotaibi, Reem, Ahlam Ali, Haya Alharthi, and Renad Almehamdi. "AI Chatbot for Tourist
Recommendations: A Case Study in the City of Jeddah, Saudi Arabia." (2020): 18-30.
IJCRT1601009 International Journal of Creative Research Thoughts (IJCRT) [Link] 34
View publication stats
