Abstract:

The ever growing diffusion of more and more powerful mobile devices, supported by suitably robust network infrastructures and efficient communication systems, is leading to the development of a multitude of cooperative applications, ranging from business networking to collaborative work and education.
This paper presents an architecture for the selection and management over time of participants to cooperative m-learning activities. Each activity is meant to be composed by tasks and two situations are considered:
1. A task can be concluded within a working session;
2. If many sessions are required and one or more students change group, the task can be entrusted to other people
Three main aspects are taken into account.
The first refers to network and software optimization: on the one hand, reliable real-time intra-group communications must be guaranteed. On the other hand, each task uses applications characterized by suitable platform requirements. In order to achieve a full cooperation within each group, applications are adapted and scaled to all the members’ devices and access technologies.
The second problem concerns educational aspects: groups and activities must be defined and managed properly over time. In particular, in the proposed architecture:
1. groups are selected also on the basis of the students’ skills and attitudes, in order to offer challenging and stimulating contests;
2. expertise and participation are constantly kept under control, so that, over time, the system can autonomously adapt groups and their evolution to the capabilities detected among participants and their actual advances. In this way, a sort of “dynamic curriculum” is defined and updated along the way, more frequently than it usually happens. In particular, such curriculum refers to competences acquired during the tasks;
The last aspect considered refers to turnover management: in case some student change group membership, working continuity and quality must be guaranteed. This is achieved by comparing the people’s curriculum mentioned above. When students change group, details stored about their background can guide teachers to define a proper turnover.
The architecture comprehends: (1) an external layer for user-system communication and application scaling; (2) a decision system which performs testing phases, defines curricula and consequently manages groups over time; (3) a database storing testing and learning material, curricula and groups.
The main contribution of the proposed approach is that groups are managed in a dynamic, self adaptive way. As a matter of fact, people’s interests and actual expertise are constantly evaluated and used to optimize his work.
In this way, the system performs a self-adaptation over-time, tailoring groups and their evolution to the capabilities of participants. This kind of management, thus, proposes cooperative educational mobile services as applications of Ambient Intelligence, being based on the main components of such environments: exchange of information between users and the system through advanced mobile devices, appropriate user-system interfaces, a pervasive communication network, data analysis and decision techniques.

Keywords:

m-learning, cooperative learning, group management, self-adaptation.