Adaptabilite d'applications mobiles dans un contexte d'apprentissage

The emergence of new communication and information technologies gives students access to a huge quantity of information at any time. Students' mobility, and their information sources are increasing. Although the low capacities of mobile devices and the low bandwidth of mobile networks, make the current paradigms not adapted to the mobile learning.; Mobile devices have certain constraints related to the size of their screen and keyboard, their computing speed, their consumption of energy, as well as to the diversity of their platform of execution. This implies difficulty to have access to specific information in a context of mobility. The next generation of mobile application intends to overcome all these limitations, while allowing users to have access to concise and adapted information, even when they are in motion.; In this thesis, we propose an innovative architecture of service allowing mobile students to receive data adapted to their geographical context, and to the characteristics of their mobile devices. From this architecture, we propose use cases that enable the detection of the user and the context of use, and also, support the access to adequate information by the students. Our model of service meets the majority of the constraints dictated by the use of mobile platforms. Detection, identification and transfer of data to mobile users are done using the Bluetooth protocol carried out on a J2ME platform. This platform is designed to work on several operating systems.; Our results showed that our model of application can be implemented on several mobile devices equipped with the execution platform Java. Moreover, when basic Bluetooth stations are placed at strategic places with high attendance, the mobile users can receive information adapted to their geographical context coming from an Internet server, while being in motion.