Keywords:Indoor stay-region, trajectory mining, location-information clustering
Location-based technology is key for ubiquitous society. Recognizing the places and their patterns where a person and an object have visited in addition to geographical locations are needed to enhance the location-based services. Although some researchers have developed methods which extract outdoor stay-regions from GPS trajectories to recognize the visiting places, there isn't technology to recognize stay-regions, such as spatial location in a living house and an office building, from indoor trajectories. Technologies for extracting an indoor stay-regions are required to achieve more intelligent indoor-location-based services, such as smart-home and smart office. Therefore, we developed a method which extracts stay regions from UWB indoor trajectory. An UWB indoor-positioning technology provides location information with a-few-tensof-centimeters error. Our developed method was evaluated comparing with conventional methods.