A Method for Recognizing Driver's Location Context with a Vehicle Information Device

Abstract

Recently, smartphones have become widely used as vehicle information devices. As a result, several advantages have appeared. First, a smartphone can log driving data such as location, vehicle status information, and so on. The logs do not depend on the car, but on the smartphone user. Second, a smartphone can gather driving-related information via the Internet. We believe that the advantages enable us to recognize several contexts. Therefore, we proposed a system that supports the user in changing provided services based on the context. We have discussed context recognition in order to create the system. One of the discussion outcomes is that the location context holds a very important role in guessing which function is required by a driver. However, it is difficult to recognize the location context using only latitude-longitude logs. In this paper, we propose a method for recognizing location context. The method is separated into three steps. First, values are recorded in a travel history such as a meshed map. The values represent driving experience points in a travelling area. Second, the method searches the travel history for a given driving area. Then, the value of the searched area is compared with a specified threshold. As a result, the location context is recognized. We implemented a prototype system and evaluated the proposed method. The results of the evaluation confirm the effectiveness of both the proposed method and expected future work.