چكيده به لاتين
Today, with the expansion of devices equipped with global positioning sensors, a new type of data, called trajectory data, has been produced, which contains very rich information about the location of objects, the way objects move, the movement pattern between objects, etc. One of the most important features hidden in trajectory data is called co-movement. Co-movement is a pattern between several objects that move together for a period of time. Co-movement problem plays an important role in many applications related to trajectory data. For example, predicting the future or compressing the trajectory data are among the applications of discovering co-motion patterns in the trajectory data. Current algorithms and solutions consider the two states of streaming or non-streaming of these data and provide online or offline models, current offline approaches assume that all the required data are available before the start of processing and work based on it, on the other hand, in the flow model, it is assumed that the data enters the system in real time, and the problem of detecting co-motion patterns is challenging. In this thesis, we intend to investigate the relationship between people in the society by discovering co-movement relationships from trajectory data and label their relationship type. In this thesis, with a new look at co-movement patterns and stepping on the latest research of scientists in this field, we try to improve the current methods of discovering co-motion and present a new concept called co-motion pattern relationship. And based on the experiments, we show that the presented method has the necessary efficiency in this field.
