Elsevier

Pattern Recognition Letters

Volume 131, March 2020, Pages 234-243
Pattern Recognition Letters

CHAT-Bot: A cultural heritage aware teller-bot for supporting touristic experiences

https://doi.org/10.1016/j.patrec.2020.01.003Get rights and content

Highlights

  • Develop a methodology in the field of tourism able to define the context based on the analysis of the text.

  • Satisfying the needs of the tourist, based on personal preferences and the current context.

  • Develop a chatbot model that uses contextual information to maintain a fluid and natural dialog with users.

Abstract

Cultural heritage is an important resource that allows us to know and promote a territory. In this respect, it is important to experiment with the enhancement of cultural heritage by adopting approaches that meet the dynamic needs of various types of users. The aim of this paper is to introduce a recommender system capable of developing adaptive tourist routes. In fact, the proposed system suggests points of interest and related services according to both the profile of the tourist and contextual aspects. In particular, the interaction of the user with the system occurs through a chatbot that allows to build a real dialog. In order to show the potential of the proposed approach, a prototype was developed to support the user in building a customized tourist route related to some of the most important cultural sites in Campania (a region in Southern Italy): Herculaneum, Paestum and Pompeii.

Introduction

The pervasiveness of digital technologies has led to the replacement of traditional data management and recommendation services with sophisticated systems. These systems can integrate data and services extracted from heterogeneous sources to provide value-added information. Nowadays a huge amount of information is used in mobile applications, even if there is the risk that this huge amount of data could generates confusion. The filtering of data and services, based on the context, arises to model possible usage scenario.

The reason for this “information deluge” is due, in large part, to the proliferation of devices that have an active role (sensors and, in general, all those objects defined as “smart”). Those devices are able to communicate with each other creating a highly pervasive network, which is at the base of the new paradigm of the Internet of Things (IoT) [1], that represents an ecosystem thanks to which “context-aware” applications exist [2], [3], [4], [5], [6]. In fact, in information management, these systems are mainly dedicated in determining which part of the entire information is relevant with respect to environmental conditions, allowing us to offer to the user a wide range of services that could help him in everyday life, work or private, in order to better manage the resources such as the time available. The parallel development of these paradigms (IoT and Context Aware) acquires fundamental importance in the design of smart environments capable of tailoring itself to users [7].

In addition, technological evolution has also been followed by a human beings’ behavior evolution: we are faced with an ever-increasing number of “digital” and “social” users, who are accustomed to use several mobile devices to search and share every type of information in multiple situation [8]. In this scenario, e-Tourism is one of the most investigated application domains [9], in which a change of context causes a transformation of the experience even before being lived [10]. Thanks to new technologies, in fact, a tourists can use several services able to filter the huge amount of data present on the network in order to only provide relevant information according to the context [11].

The aim of this paper is to introduce a chatbot based on a Context-Aware System able to recommend contents and services according to tourist profiles and context. This chatbot is capable to maintain, through the combination of Pattern Recognition and context recognition techniques, a logical conversation with the user and to supply to specific tourist needs. The proposed architecture would be able to control the evolution and presentation of information to the user based on different types of context. The chatbot could be seen as a modern tourist guide, which allows the dynamic delivery of several information, services, or narrative content (textual and multimedia) properly integrated and tailored to the needs and dynamic behavior of users.

Section snippets

Background

The concept of context does not have a rigorous definition, and, over the years, many different interpretations of its meaning have been given. In fact, this concept plays an important role in many different disciplines, such as psychology, linguistics and computer science, and in each of these can take on a different meaning, more fitting for its application.

Related works

The importance of context in information technology has increased in recent decades as computers have become increasingly pervasive in everyday life. Context awareness, or the idea that these systems can detect and react to a user situation, is a popular research topic. While the computer community has initially considered the context as a question of user position, in the last few years this concept has been considered as part of the whole process in which users are involved. Sophisticated and

The proposed architecture

The architecture of the CHAT-Bot is based on some main modules, as shown in Fig. 1. The storytelling module is closely related to the bot ability to guide the user through the whole experience, making the way of proceeding while leaving the user free to express himself and immerse himself in the personalized story of a place.

Each user acts differently from the others, becoming part of a creative process and creating a unique and unrepeatable visit. In planning an itinerary and in designing the

Experimental phase

Based on the proposed architecture, an application prototype was developed: a chatbot, designed and implemented, along with a server-side component, as described above (Fig. 4). The chatbot was initially designed to support tourists visiting some of the cultural sites of the Campania region of Italy (Paestum, Pompeii, Herculaneum). In this experimental phase, the main services and contents potentially useful for tourists have been identified.

After the interaction with the chatbot, 3150 users

Conclusions

Today, the amount of data and services available require not only their mutual integration, but also their filtering in order to: provide the user in an appropriate way with a set of tailored data and services; operate on a manageable amount of data to improve processing efficiency; provide the user with only what is relevant based on contextual aspects, such as location and time.

In the world of tourism, for example, all this can be used to deepen the logic of integration and interoperability

Declaration of Competing Interest

None.

References (29)

  • F. Amato et al.

    Recognizing human behaviours in online social networks

    Comput. Secur.

    (2018)
  • F. Colace et al.

    Revising recurrent neural networks from a granular perspective,

    Appl. Soft Comput.

    (2019)
  • K. Ashton

    That ‘internet of things’ thing,

    RFID J.

    (2009)
  • D. Capriglione et al.

    Estimating the Outdoor PM10 Concentration Through Wireless Sensor Network for Smart Metering,

    (2019)
  • F. Colace, D. Santaniello, M. Casillo, and F. Clarizia, “BeCAMS: A behaviour context aware monitoring system,” in 2017...
  • F. Clarizia et al.

    A multilevel graph approach for rainfall forecasting: A preliminary study case on London area,

    Concurr. Comput. Pract. Exp.

    (2019)
  • P. Albano et al.

    Secure and distributed video surveillance via portable devices

    J. Ambient Intell. Humaniz. Comput.

    (2014)
  • F. Colace, M. Lombardi, F. Pascale, D. Santaniello, A. Tucker, and P. Villani, “MuG : A Multilevel Graph Representation...
  • G. Annunziata et al.

    Appoggiomarino: A context Aware app for e-citizenship,

  • G. D'aniello, M. Gaeta, and M.Z. Reformat, “Collective perception in smart tourism destinations with rough sets,” in...
  • M. Casillo et al.

    An approach for recommending contextualized services in e-tourism,

    Information

    (May 2019)
  • F. Amato et al.

    Big data meets digital cultural heritage

    J. Comput. Cult. Herit.

    (2017)
  • B.N. Schilit et al.

    Disseminating active map information to mobile hosts,

    IEEE Netw.

    (1994)
  • J. Pascoe et al.

    Issues in developing context-aware computing,

    Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

    (1999)
  • Cited by (48)

    • Where do I belong? A study of associations between guanxi capital and local identity through WeChat use among Chinese youth

      2023, Technological Forecasting and Social Change
      Citation Excerpt :

      As local attachment and local identity belong to the sub-dimension of a local sense of belonging and are indicators to measure the emotional intensity of young residents, the guanxi capital of youth can predict local identity. Additionally, Casillo et al. (2020) claim that humans are faced with an ever-increasing number of ‘digital’ and ‘social’ users, who are accustomed to using several mobile devices to search for and share every type of information in multiple situations; in this scenario, a change of context causes a transformation of the experience even before it is realised. To summarise, media use can significantly predict the local identity of a new local population.

    • Personalized recommendation of collective points-of-interest with preference and context awareness

      2022, Pattern Recognition Letters
      Citation Excerpt :

      Amato et al. [18] introduced a methodology integrating recommendation facilities with agent-based planning techniques in order to implement a planner of routes within cultural sites as museums. Casillo et al. [19] developed a recommender system that suggests points of interest and related services according to both the profile of the tourist and contextual aspects. Unfortunately, the current works, including the above mentioned methods, usually fail to learn the abstract interaction among the features of users, POIs and time, or do not provide the solution to the problem of collective POIs recommendation as our work addresses.

    • A cultural heritage framework using a Deep Learning based Chatbot for supporting tourist journey

      2021, Expert Systems with Applications
      Citation Excerpt :

      In particular, Miah et al. (2017) and Valls et al. (2018) aim to cluster travelers for management operation instead of supporting users in their journey while Lee et al. (2019b) and Pallud (2017) design an immersive and/or interactive experience for a specific cultural object; Castiglione et al. (2018), Machidon et al. (2020) and Moustaka et al. (2019) are aimed at information processing whilst (Casillo et al., 2020) support tourists through a story telling experience profiling them according to information made in registration phase; Castiglione et al. (2018), Machidon et al. (2020), Moustaka et al. (2019) and Dou et al. (2018) provide only user and organization oriented services respectively.

    • Adaptive and hybrid context-aware fine-grained word sense disambiguation in topic modeling based document representation

      2021, Information Processing and Management
      Citation Excerpt :

      We evaluate the method in Section 4 and draw conclusions in Section 5. Document representation has long been studied in various areas (Le & Mikolov, 2014) and widely used in real-life applications, e.g., public sentiment analysis (Pandey, Rajpoot, & Saraswat, 2017), chat-bot (Clarizia et al., 2018) and e-learning (Casillo et al., 2020). Topic modeling and word embedding are two important paradigms for this task.

    • Critically assessing AI/ML for cultural heritage: Potentials and challenges

      2023, Handbook of Critical Studies of Artificial Intelligence
    View all citing articles on Scopus
    View full text