Context and Sentiment Aware Chatbot for Tourism Applications.

Recently, advances in the informatics field have changed the tourism and hospitality sectors drastically. The interaction between the service provider and the customer is revolutionized by employing sophisticated chatbots. Although there are a lot of chatbots, most of them are not sentiment and context-aware. In this study, we developed a context and sentiment-aware chatbot for tourism applications. First, we created a database that includes frequently asked questions in the tourism area. We then automatically correct mistakes in the customer queries and compute the similarity of queries with existing sentences in the database and respond to them with the corresponding answer. We also monitor the sentiment of users continuously and try to estimate the context (before sales and after-sales). The proposed system has promising results for real-life applications.