Toward a realization of similarity search for moving object databases

摘要

We aim for constructing a moving object database system that enables us to store moving objects in the real world and issue queries for various types of object moves. In general, the words in a natural language such as "rotate" or "wave" are difficult to express the exact move of an object that a user wants to specify. So, it seems natural to introduce a query-by-example (QBE)-based similarity search on object move in a moving object database system. In this paper, we first investigate a various types of the identity and the similarity of object moves, and then introduce a moving object query model which takes into account not only the time and location of moving objects but also the time warping and the spatial transformation of object move so that we can answer to a variety of spatio-temporal queries about object move. A fundamental experience about similarity the QBE-based moving object search is also reported in this paper.