Point of interest data storage using ontology

With the rapid development of urbanization in the world, information is continuously exploding all the time. Via portable devices such as smart phones, a large amount of information is generated and multicast. The interactions among end-users build social networks, such as Twitter, Foursquare and alike, which enable people to exchange tweets and pictures. Social network information is commonly linked to Points of Interest (POIs), which are essentially physical locations. Therefore, POIs are the key to identify information on accidents, events, meetings, etc. However, similar POI information may be provided by different people or networks. Without a unified structure, such information may result in uncontrollable redundancy. And sharing information would be more difficult, thus contrasting the assumption that sharing data is one critical achievement of the social networks. Many solutions could be used to provide services on collecting and matching data from different platforms and identifying redundant information (i.e. different messages describing the same issue). Some solutions use either clustering or data ranking algorithms to match different data, while others may use open linked data (semantic Web) to store POIs, thus making it easier to match data. However, without a unified structure or query approach, the problem could not be really solved. In this paper, we contribute a novel approach for storing POI data by using ontology, which is capable of building unified data structures and integrating data, as well as providing a unified query approach.