On-site Likelihood Identification of Tweets for Tourism Information Analysis

Tourism is one of the most important key industries. The Web contains much information for the tourism, such as impressions and sentiments about sightseeing areas. Analyzing the information is a significant task for tourism informatics. One approach to extract tourism information is to extract sentences with keywords related to target facilities and events. However, all sentences with keywords might be not tourism information. In this paper, we propose a method for measuring tourism information likelihood. The target resource for the analysis is information on Twitter. The task is to identify whether each tweet has high on-site likelihood. We introduce a filtering process and a machine learning technique for the task. Our method obtained 80.5% on the precision rate.