Building an earthquake evacuation ontology from twitter

During the massive Tohoku earthquake, while landlines and mobile phone lines got stuck, Twitter were used to exchange information about evacuation. On 11 March, the number of tweets from Japan dramatically increased to about 33 million, 1.8 times higher than the average figure. However, since texts on Twitter are unstructured data, are more complex than other text media, it is difficult to find a suitable evacuation center. Not only information in Twitter, people but also need to combine other information such as map information, shop information to find the suitable evacuation center. Since it takes much time to do these tasks, these manual processing are not suitable in the emergency status. Therefore, we need an approach to help computers to understand the meaning of evacuation, and to provide the most suitable evacuation center based on earthquake victims' behaviors in real-time. In this paper, we firstly design an earthquake evacuation ontology. Secondly, we indicate that by using this ontology, computers can provide the most suitable evacuation center based on earthquake victims' behaviors in real-time.