A tourist advisor based on a question answering system

Abstract

Tourism is one of the world's largest industries with a global economic contribution in the order of billions of dollars each year. With the intensified competition in the sector, it has become paramount to facilitate the creation of highly personalised services and experiences. In order to achieve this, most tourists currently rely on search engines when seeking relevant information. In this paper, we introduce a Tourist Advisor which uses a Question Answering System (QAS) to provide accurate answers to the tourists' questions instead of simply identifying a set of documents that are likely to contain some relevant information. The QAS has several innovative features, including the utilisation of a “controlled” natural language as a means of representation of both the domain knowledge and the common sense. We present the architecture of the system and discuss its deployment in a pilot study carried out in one of the most renowned Italian tourist destinations. Results show that when compared with traditional web search, our system provides much concise and precise answers in a shorter amount of time.