A method of estimating outdoor situation for lifelog generation

We propose a method of estimating the situation in the outdoors for generating high quality lifelog. The algorithm of estimating outdoor situation is possible to distinguish and estimate several situations in the same location. This algorithm estimates situations by using a decision tree of machine learning. We do decision tree analysis by using explanatory variables such as transportation means, sensor data, and social data. Experimental results show that we can achieve high levels of accuracy which F-Measure is about 0.8 or more.