Exploring conversational assistants to provide accessibility information in tourism

• Autor(a) principal: Correia, Tiago Jorge Costa
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/33880
• topic: Conversational assistant; RASA; Accessible tourism; Conversation-driven development

Although nowadays tourism is one of the most important economic activities worldwide, contributing 10.2% of the Gross Domestic Product (GDP) and 10% of the total employment generated, there are still several groups in our society that face many travel constraints, e.g., persons with disabilities (PwD). The contribution of tourism to the promotion of an inclusive society implies the adoption of strategies that promote the accessibility of tourism destinations, allowing all people, regardless of their abilities, to enjoy touristic experiences. Information on the accessibility of different spaces and infrastructures plays a significant role in how people with different types and levels of disability can plan their lives, especially their touristic experiences. In this regard, important challenges exist, not only in gathering all the relevant information, but also, and most of all, making it available to PwD in ways that are also accessible and adaptable to different needs and abilities. The failure to present this information usually leads people to give up living tourist experiences because they do not feel comfortable to do it in an environment that may not be adapted to their needs and concerns. Although it already exists, the amount of information makes it complex to reach the information in a simple and intuitive way. Therefore, having a more simple and natural way to get to the information that you want is an important aspect to explore. In this context, supporting dialogue in natural language, whether by voice or writing, can bring advantages as a complement or an alternative to the more common interfaces, such as portals or menus, to present the information, as it can potentially bring some of the naturalness and efficiency of speech communication to the interaction with these systems. And while it can be a relevant interaction alternative for all, it can favor those that might face difficulties in accessing and learning more traditional systems. In this work we present the results for the design, development and instantiation of a conversational assistant for Accessible Tourism. Based on a conversation-driven development methodology and considering RASA for the development and RASA X backend for the deployment of the assistant, the use of the assistant by domain experts enabled an iterative process for its refinement and improvement. At its current stage of development, the assistant enables querying about hotels, museums and locals of interest suggestions along with related accessibility information and provides promising grounds for further evolutions.