<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Intelligent Smart Tourism Education: AI-based Learning for Cultural Tourism Experiments</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Michele Angelaccio</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Michele Fasolo</string-name>
          <xref ref-type="aff" rid="aff0">0</xref>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Lucia Zappitelli</string-name>
          <email>lucia.zappitelli@gmail.it</email>
          <xref ref-type="aff" rid="aff0">0</xref>
          <xref ref-type="aff" rid="aff1">1</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Dipartimento Ingegneria dell'Impresa (DII), University of Rome “Tor Vergata”</institution>
          ,
          <addr-line>via del Politecnico 1, Rome, 00100</addr-line>
          ,
          <country country="IT">Italy</country>
        </aff>
        <aff id="aff1">
          <label>1</label>
          <institution>Ital-IA 2024: 4th National Conference on Artificial Intelligence</institution>
          ,
          <addr-line>organized by CINI</addr-line>
        </aff>
      </contrib-group>
      <abstract>
        <p>Self-learning and active learning are becoming key focal points in the digital era, where the demand for online learning is growing. This is particularly important for the education of smart tourism guides within the context of cultural tourism. In this paper, we present some experiences obtained within the context of the Digital Tourism Course at the University of Rome Tor Vergata over the past two years. The focus is on enhancing self-learning skills through the use of AI generative techniques and its role in digitalizing cultural tourism. We also provide comparisons and discussions on the reported advantages.</p>
      </abstract>
      <kwd-group>
        <kwd>eol&gt;Smart cultural tourism</kwd>
        <kwd>inclusive learning</kwd>
        <kwd>self-learning</kwd>
        <kwd>AI</kwd>
        <kwd>generative models1</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>1. Introduction</title>
      <p>The integration of artificial intelligence (AI) in
education, particularly in the field of digital tourism, is
a growing area of interest. Xing (2022) and Cesta
(2020) both emphasize the potential of AI in
personalizing learning experiences, with Xing
focusing on the design of a tourism teaching system
and Cesta on the use of intelligent tools for cultural
heritage visits. Ferràs (2020) further explores the
application of AI in tourism, highlighting its role in
creating customized experiences. Morellato (2014)
adds a new perspective, proposing an experiential
approach to developing digital competence in tourism
education, which could be enhanced by AI-driven
personalization and active learning. These studies
collectively underscore the potential of AI in
transforming the learning experience in digital
tourism, from personalized teaching systems to the
creation of tailored tourist experiences.</p>
      <p>In recent years, the integration of artificial
intelligence (AI) into education has been rapidly
advancing, offering promising opportunities for
enhancing learning experiences. One of the notable
applications of AI in education is AI-assisted learning,
which leverages machine learning algorithms and
natural language processing techniques to
personalize and optimize the learning process for
individual students. This introduction aims to provide
an overview of AI-assisted learning, drawing insights
from a survey of relevant literature.</p>
      <p>
        AI-assisted learning encompasses a variety of
techniques and technologies aimed at tailoring
educational content, delivery, and assessment to meet
the diverse needs and preferences of learners. As
highlighted by [
        <xref ref-type="bibr" rid="ref4">4</xref>
        ], AI systems can analyze vast
amounts of educational data, including student
performance, learning styles, and knowledge gaps, to
generate personalized learning pathways and
recommendations. These systems can adaptively
adjust the difficulty level of learning materials,
provide real-time feedback, and offer additional
resources or exercises based on individual learning
progress and proficiency.
      </p>
      <p>
        Moreover, AI-powered tutoring systems have
shown great potential in providing personalized
© 2024 Copyright for this paper by its authors. Use permitted under
Creative Commons License Attribution 4.0 International (CC BY 4.0).
support and guidance to students. For instance,
intelligent tutoring systems (ITS) can simulate
oneon-one interactions with students, offering immediate
feedback, explanations, and hints tailored to their
specific learning needs [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]. Through the use of natural
language understanding and dialogue generation
techniques, chatbot-based tutoring systems can
engage in conversational interactions with learners,
answering questions, clarifying concepts, and
fostering a supportive learning environment [
        <xref ref-type="bibr" rid="ref5">5</xref>
        ].
      </p>
      <p>
        Furthermore, AI-assisted learning extends beyond
individualized tutoring to collaborative and social
learning experiences. Social recommender systems
can leverage AI algorithms to analyze learners' social
interactions, preferences, and interests, facilitating
the discovery of relevant learning resources and
fostering peer-to-peer knowledge sharing [
        <xref ref-type="bibr" rid="ref6">6</xref>
        ]. Virtual
collaborative environments powered by AI can enable
students to engage in collaborative problem-solving,
project-based learning, and group discussions, with
intelligent agents providing guidance, coordination,
and feedback throughout the collaborative process
[
        <xref ref-type="bibr" rid="ref3">3</xref>
        ]. In this paper we outline some experiences
obtained in the context of Digital Tourism Course at
the University of Rome Tor Vergata in the last 2 years.
The focus is on improving self-learning skills with the
use of AI generative and its role in the context of
cultural tourism digitalization motivated by several
reasons as described in [
        <xref ref-type="bibr" rid="ref1">1</xref>
        ] in which the importance of
academic role has been outlined.
2. Impact of AI in Providing
Self
      </p>
      <p>Learning Activities
The impact of AI in providing self-learning activities is
profound and multifaceted. Here are several key ways
in which AI influences self-learning:
• Personalized Learning Paths
AI can analyze vast amounts of data about a
learner's preferences, strengths, weaknesses, and
learning styles to tailor educational content and
activities to their individual needs. By
understanding the learner's pace and level of
understanding, AI algorithms can suggest
appropriate resources, exercises, and challenges,
thereby enabling self-directed learning that is
customized to each learner.
• Adaptive Learning Systems
AI-powered adaptive learning systems
dynamically adjust the difficulty and pace of
learning activities based on the learner's
performance and progress. These systems can
provide personalized recommendations for
activities that match the learner's current level of
knowledge, ensuring that they are appropriately
challenged without feeling overwhelmed or
bored. By adapting to the learner's evolving
abilities, adaptive learning systems promote
continuous improvement and engagement in
selflearning activities.
•
•
•</p>
      <p>Intelligent Tutoring Systems (ITS)
ITS leverage AI techniques to simulate
oneon-one tutoring interactions, providing
personalized guidance, feedback, and
support to learners. Through natural
language processing and machine learning
algorithms, ITS can understand the learner's
questions, misconceptions, and learning
goals, offering targeted explanations, hints,
and examples to facilitate self-directed
learning. By providing immediate and
tailored assistance, ITS empower learners to
navigate complex concepts and topics
independently.</p>
      <p>Recommendation Systems
AI-driven recommendation systems analyze
learners' preferences, past interactions, and
learning objectives to suggest relevant
resources, courses, and activities for
selfdirected learning. These systems can
recommend educational materials, such as
articles, videos, tutorials, and online courses,
that align with the learner's interests and
goals, thereby facilitating serendipitous
discovery and exploration of new topics. By
curating personalized learning pathways,
recommendation systems empower
learners to take ownership of their learning
journey and pursue areas of interest
autonomously.</p>
      <p>Natural Language Processing (NLP)
NLP technologies enable interactive and
conversational interfaces for self-learning
activities, such as chatbots and virtual
assistants. Learners can engage in dialogue
with AI-powered agents to ask questions,
seek explanations, and receive feedback in
natural language. By fostering interactive
and responsive learning experiences,
NLPdriven systems support self-directed
inquiry, reflection, and problem-solving,
enhancing learners' autonomy and
confidence in their ability to learn
independently.</p>
      <p>Overall, AI plays a transformative role in enabling and
enhancing self-learning activities by providing
personalized guidance, adaptive support, curated
resources, interactive interfaces, and actionable
feedback. By leveraging the capabilities of AI
technologies, learners can engage in self-directed
learning that is tailored to their individual needs,
preferences, and aspirations, thereby empowering
them to become lifelong learners mastering new
skills, acquiring knowledge, and achieving their
educational goals autonomously.
3. Case Description: Self-Learning in</p>
      <p>Digitalization of Cultural Tourism
Digitalization of cultural routes is a key point used in
the education of new generation of tour operators and
smart guides. Self-Learning AI powered activities
provide in this case a promising means to guarantee a
flexible and dynamic learning process suitable in the
new tourism era and in particular cultural tourism
case. To demonstrate the applicability of AI based
education in cultural tourism, we consider three
examples used in the Rome Tor Vergata University.
3.1. Self-learning Platform for Cultural</p>
      <p>Travel Blog Design</p>
      <p>A first Example of Digital Tourism learning activity
is the digitalization of cultural routes emerging from
the need to explore archeo sites and cultural heritage
assets in a huge number of ways. As example of
selflearning platform, we discuss the case of cultural
travel blog designed for walking routes and denoted
as cultural travel blog in which information is
designed to offer comprehensive resources and
interactive activities focused on the digitalization of
cultural tourism. Learners have access to a variety of
multimedia content, including articles, videos, case
studies, and tutorials, covering topics such as digital
marketing strategies for cultural attractions, virtual
reality experiences in heritage sites, and augmented
reality applications for guided tours.</p>
      <sec id="sec-1-1">
        <title>3.1.1. Personalized Learning Paths</title>
        <p>Upon accessing the platform, learners are
prompted to create a profile where they can specify
their interests, goals, and prior knowledge in the field
of cultural tourism. AI algorithms analyze this
information to generate personalized learning paths
tailored to each learner's preferences and learning
objectives. For example, learners interested in digital
marketing may receive recommendations for articles
and tutorials on social media advertising for cultural
destinations, while those interested in immersive
experiences may be directed to resources on virtual
reality technologies.</p>
      </sec>
      <sec id="sec-1-2">
        <title>3.1.2. Chatbot Assistance</title>
        <p>To facilitate seamless navigation and access to
resources, the platform features a chatbot interface
that serves as a virtual assistant. Learners can interact
with the chatbot using natural language queries to
request specific information or resources. For
instance, a learner interested in learning about the use
of augmented reality in cultural tourism may type "AR
applications" into the chatbot interface. The chatbot
then retrieves relevant PDF documents, articles, and
videos from the platform's database and presents
them to the learner for exploration.</p>
      </sec>
      <sec id="sec-1-3">
        <title>3.1.3. Accessing PDF Resources</title>
        <p>One of the key functionalities of the chatbot is its
ability to provide access to PDF resources on demand.
Learners can request PDF documents on specific
topics by simply typing keywords or phrases related
to their interests. For example, a learner curious about
the impact of digitalization on cultural heritage
preservation may type "heritage preservation" into
the chatbot interface. The chatbot then searches the
platform's repository for PDF documents, reports, or
research papers related to the topic and presents
them as downloadable resources.</p>
      </sec>
      <sec id="sec-1-4">
        <title>3.1.4. Interactive Learning Activities</title>
        <p>In addition to accessing static resources, learners
can engage in interactive learning activities to deepen
their understanding of digitalization in cultural
tourism. These activities may include quizzes,
simulations, virtual tours, and collaborative projects,
allowing learners to apply theoretical concepts in
real-world scenarios and gain hands-on experience in
digital cultural heritage management.</p>
      </sec>
      <sec id="sec-1-5">
        <title>3.1.5. Progress Tracking and Feedback</title>
        <p>Throughout their learning journey, learners can
track their progress and performance using built-in
analytics and assessment tools. AI algorithms analyze
learner interactions, quiz scores, and completion rates
to provide personalized feedback and
recommendations for further learning. Learners
receive insights into their strengths and areas for
improvement, empowering them to adapt their
learning strategies and goals accordingly.</p>
        <p>BEFOREMaster
AFTER Master
++AIeepcnoognwaonemgreeicmdbesenkntilesfits DesTtPirnaeLaioontpcieoladenl OSnelinlfelelaeranrinnigng</p>
        <p>SmartTourOperator</p>
        <p>Skils
Destination LOCPArLacktnicoawllSekdiglesand</p>
        <p>Local
People</p>
        <p>Tour
Operator
Focus</p>
        <p>Marketing,Communication</p>
        <p>TRAVELSkils</p>
        <p>Destination
Planning Designer</p>
        <p>Technical,
Experience
TRAVELSkils</p>
        <p>Cultural</p>
        <p>Expert
Colaborate</p>
        <p>Nar ative
CULTURAL</p>
        <p>Skils
AI
Models</p>
        <p>Travel
Guide</p>
        <p>Destination
LESSPlanning Designer</p>
        <p>AITravel
chatBot
AI Travel Guide</p>
        <p>Master</p>
        <p>Organization
AI Builder</p>
        <p>AIgenerated
Packet</p>
        <p>Cultural
Expert
ExpeArtI</p>
        <p>The self-learning platform for digitalization of
cultural tourism offers a flexible and engaging
educational experience, enabling individuals to
acquire knowledge and skills at their own pace and
convenience. By harnessing the capabilities of AI
technologies and interactive resources, the platform
empowers learners to explore the intersection of
culture, technology, and tourism and become
informed advocates for sustainable and inclusive
digital practices in cultural heritage preservation and
promotion.
3.2. Self-learning activities for local Smart</p>
        <p>Cultural Operators</p>
        <p>As another example of the application of a
selflearning platform, we highlight the case of Archeo
Tutorial, implemented within the context of the
ERASMUS+ project ADHOC. The platform's role is to
enhance accessibility for users with disabilities. In this
context, the platform has been equipped with a
dynamic generative script for the text-to-speech
function, allowing it to be seamlessly integrated into
the original archaeological guide. Obviously, this
generation can be put in the form of NLP by taking
advantage from a chatbot interacting with PDF
resources. This approach showcases how AI
generative interaction with PDF resources can serve
as a viable solution for self-learning activities aimed at
educating cultural guides. This is particularly valuable
in contexts where there is a need to extend cultural
tourism learning to students with disabilities, as part
of an inclusive learning program. For instance, in the
case of translating cultural content into LIS (Italian
Sign Language), the learning program is tailored for
students who are already proficient in using LIS.</p>
        <p>Figure 1 illustrates the impact of AI-powered
learning methodologies on the education of local
cultural guides. The diagram depicts the gains and
benefits that could be achieved through the Master's
program (see also 8]).</p>
        <p>In this latter example, we aim to highlight an
instance of an AI-powered application utilizing
advanced AI-assisted self-cultural guidance. The
learner communicates their interest in cultural routes
in a densely populated cultural destination such as
Rome through a natural language query to the
chatbot. The chatbot recognizes the learner's request
and responds by asking a PDF book.</p>
        <p>Exploring historical cities like Rome can now
benefit from a variety of digital tools (applications,
augmented reality, chatbots, and interactive maps)
that facilitate immediate access to the dense and
hyper-dimensional fabric of data, stories, and paths
related to the cultural components of the urban
landscape. Particularly, the integration of intelligent
chatbots and interactive maps, programmed with
extensive databases of information, represents an
innovative breakthrough in the cultural tourism
sector capable of offering users immediate answers to
specific questions, recommendations tailored to their
interests and needs, as well as real-time updates on
events in the explored context, significantly enriching
the personal experience and making it unique and
rich. In this regard, a case study aimed at optimizing
the touristic exploration of Rome is proposed, based
on the innovative integration of an advanced chatbot,
built on generative artificial intelligence technologies,
and Leaflet, an open-source platform for creating
interactive maps. This synergy specifically aims to
offer a highly personalized tourist experience, within
the scope of themes chosen by the user, encompassing
a wide spectrum of cultural components of the urban
landscape, from artistic and historical to
enogastronomic elements. This system, named
GuidaTuristicaAI (or AI-CH-Tour-Map), includes in its
descriptions information drawn from a vast corpus of
documents, ranging from books by illustrious visitors,
literary works, and musical compositions, to
enogastronomic guides about Rome over the
centuries. Upon interacting with GuidaTuristicaAI, the
user is invited to select one or more themes of interest
to further personalize their exploration of the city. If
the choice falls, for example, on literary works about
Rome as a thematic filter, the chatbot searches its PDF
document database to select those including literary
references to Rome, extracted from works by famous
authors who have described the city in their visits or
narrative and poetic works. In the case study, it is
imagined that an art history student plans a visit to
Rome following the footsteps of Giuseppe Vasi. The
chatbot suggests a series of publications. For example,
they can download the PDF of one of the volumes from
the open archive.org site or another
source:( https://ia801006.us.archive.org/7/items/te
sorosacroevene02vasi/tesorosacroevene02vasi.pdf )
. Using the web application, the user receives a
personalized itinerary with the sites described by
Vasi. The generated interactive map allows him to
easily navigate from one place to another, while the
chatbot provides cultural and historical insights. The
Leaflet mapping platform plays a crucial role in the
implementation of this thematic customization. The
dynamism of the real-time interaction between the
user and GuidaTuristicaAI can indeed be enriched
with specific points of interest identified on the map
related to other themes selected by the user, allowing
them to discover the city through different
interpretative and perceptual lenses. For example, in
accord to the literary theme, markers can be added to
places mentioned in novels or poems, with short
excerpts or quotes visible by clicking on the respective
markers, emotionally connecting them to the places
through the words of authors who have immortalized
them in their works. Moreover, information can be
provided on dining venues and typical Roman dishes
and wines that can be paired with them.</p>
      </sec>
    </sec>
    <sec id="sec-2">
      <title>4. Conclusions</title>
      <p>From the outset and in the coming years, AI
promises to play a decisive innovative role in the field
of educational pathways, transforming and enhancing
learning experiences through its technological
features (intelligent tutoring systems and chatbots)
that cater to the needs for personalization and
adaptability. This work particularly investigates the
importance and impact of AI, with a special reference
to AI-assisted learning in the field of cultural tourism.
There are several ways in which AI can be enabled for
better personalization and improvement of the
educational experience, namely customizing learning
paths, advanced tutoring systems, recommendation
mechanisms, and the use of natural language
processing to enhance interaction and support for
students. The research has specifically focused on the
use of self-learning platforms in the production of
blogs dedicated to cultural travel, emphasizing the
importance of digitalizing cultural routes. Particular
attention was given to specific use cases implemented
at the University of Rome "Tor Vergata", highlighting
how AI solutions can enable self-learning and
accessibility for users with disabilities, aiming to
ensure a better experience in their training for future
professionals. AI can help and enhance a series of
activities conducted in self-learning through personal
guidance, adaptive support, and interactive interfaces
to resources. It allows students to follow a
selfregulated learning path that, according to their needs,
preferences, and personal goals, enables approaches
to lifelong learning, which can empower them to
acquire new skills and knowledge entirely
autonomously. Such flexible and interactive
selflearning modalities can be effectively extended into
digital tourism, facilitating experiences by any user in
consuming cultural landscapes in a more accessible
and engaging way.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <surname>Amato</surname>
          </string-name>
          ,
          <string-name>
            <surname>Flora</surname>
          </string-name>
          , et al.
          <article-title>"Ai-powered learning: Personalizing education for each student</article-title>
          .
          <source>" 3rd CINI National Conference on Artificial Intelligence (ITAL-IA</source>
          <year>2023</year>
          ), in Pisa, Italy.
          <year>2023</year>
          .
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <given-names>J. R.</given-names>
            <surname>Anderson</surname>
          </string-name>
          et al. “
          <article-title>Cognitive tutors: Lessons learned”</article-title>
          ,
          <source>Journal of the Learning Sciences (1985)</source>
          , pp.
          <fpage>167</fpage>
          -
          <lpage>207</lpage>
          . doi:
          <volume>10</volume>
          .1207/s15327809jls0402_
          <fpage>2</fpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [3]
          <string-name>
            <surname>T. De Jong</surname>
            and
            <given-names>A. W.</given-names>
          </string-name>
          <string-name>
            <surname>Lazonder</surname>
          </string-name>
          . “
          <article-title>Designing Virtual Collaborative Learning Environments”</article-title>
          .
          <source>In: Learning, Design, and Technology: An International Compendium of Theory, Research</source>
          , Practice, and Policy. Ed. by
          <string-name>
            <given-names>M.</given-names>
            <surname>Spector</surname>
          </string-name>
          ,
          <string-name>
            <given-names>B. B.</given-names>
            <surname>Lockee</surname>
          </string-name>
          , and
          <string-name>
            <given-names>M. D.</given-names>
            <surname>Childress</surname>
          </string-name>
          . Springer,
          <year>2016</year>
          , pp.
          <fpage>1</fpage>
          -
          <lpage>23</lpage>
          . doi:
          <volume>10</volume>
          .1007/978-3-
          <fpage>319</fpage>
          -17727-4_
          <fpage>179</fpage>
          -
          <lpage>1</lpage>
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [4]
          <string-name>
            <given-names>A.</given-names>
            <surname>Smith</surname>
          </string-name>
          . “
          <article-title>AI in Education: Promises and Pitfalls”</article-title>
          .
          <source>In: Journal of Educational Technology 45.3</source>
          (
          <issue>2020</issue>
          ), pp.
          <fpage>321</fpage>
          -
          <lpage>335</lpage>
          . doi:
          <volume>10</volume>
          .1080/0000000X.
          <year>2020</year>
          .
          <volume>1234567</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          [5]
          <string-name>
            <given-names>X.</given-names>
            <surname>Wang</surname>
          </string-name>
          et al. “
          <article-title>Designing chatbots for tutoring: A collaborative approach”</article-title>
          .
          <source>In: International Journal of Artificial Intelligence in Education 29.4</source>
          (
          <issue>2019</issue>
          ), pp.
          <fpage>533</fpage>
          -
          <lpage>574</lpage>
          . doi:
          <volume>10</volume>
          .1007/s40593- 019-00192-y.
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [6]
          <string-name>
            <given-names>J.</given-names>
            <surname>Yu</surname>
          </string-name>
          et al. “
          <article-title>Social recommendation for e-learning systems: A survey of the state-of-the-art and future research directions”</article-title>
          .
          <source>In: IEEE Transactions on Emerging Topics in Computing 5.2</source>
          (
          <issue>2017</issue>
          ), pp.
          <fpage>208</fpage>
          -
          <lpage>219</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          [7] https://projectadhoc.files.
          <source>wordpress.com/202</source>
          <volume>2</volume>
          /12/ad-hoc-italian.pdf
        </mixed-citation>
      </ref>
      <ref id="ref8">
        <mixed-citation>
          [8]
          <string-name>
            <given-names>Angelaccio</given-names>
            <surname>Michele</surname>
          </string-name>
          , Zappitelli Lucia, “
          <article-title>Active Learning in Teaching Digital Tourism: preliminary results through Online Travel Business Simulation</article-title>
          .”
          <source>In: eLmL</source>
          <year>2023</year>
          , The Fifteenth International Conference on Mobile, Hybrid, and
          <article-title>On-line Learning</article-title>
          . IARIA,
          <year>2023</year>
          . p.
          <fpage>1</fpage>
          -
          <lpage>4</lpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>