| Summary: | Day by day people need to move between many places, making multiple movement transitions. This type of transition involves a transportation mode, allowing people to travel, such as between home and work, work and shopping center, among other instances. Each of these activities is comprised by a set of activities, each being characterized by a transportation mode. The ability to gather such activities or transportation modes allows creation of knowledge for various applications. This dissertation seeks to contribute in solving the problem of infering transportation modes in real time on a mobile device. The developed solution will be ready to be added to the application SenseMyCity, an application developed within the Future Cities european project, as an additional module, only using the GPS sensor to gather data. It is expected to infer the main modes of transportation used, such as stationary mode, walking, car and bus. This dissertation proposes the creation of a robust and accurate computational method for infering transportation modes in two phases of classification. It is expected that this computational method can be tested using data sets from Future Cities and also public data sets.
|
|---|