A Study of Probabilistic Position Information using Absolute and Relative Position Information
暂无分享,去创建一个
Position information is classified into 2 categories: absolute position information and relative position information. The former is expressed by an absolute coordinate system such as latitude/longitude and address. The latter is expressed by a relative coordinate system which shows a positional relation of 2 objects. Absolute/relative position information differs on a coordinate system, thus they cannot cross-reference. If there is position information that can be converted from both of them to express position more flexible, it can infer new position information by a cooperation of position information expressed by different systems. In this study, we propose probabilistic position information which expresses position by combinations of area and probability of presence. The probabilistic position information can express position which is inferred from logs of absolute/relative position information and which was invisible so far. Then, we consider an algorithm that infers the probabilistic position information from 2 kinds of position information; absolute position information that expresses geographic coordinates (latitude and longitude), and relative position information that expresses pass-by of 2 objects. Finally, we consider designing of a service which applies the algorithm.
[1] Masahide Nakamura,et al. KULOCS: unified locating service for efficient development of location-based applications , 2016, Int. J. Pervasive Comput. Commun..