1: \begin{abstract}

2: We study a class of discrete-time multi-agent systems modelling

3: opinion dynamics with decaying confidence. We consider a network

4: of agents where each agent has an opinion. At each time step, the

5: agents exchange their opinion with their neighbors and update it

6: by taking into account only the opinions that

7: differ from their own less than some confidence bound. This

8: confidence bound is decaying: an agent gives repetitively

9: confidence only to its neighbors that approach sufficiently fast

10: its opinion. Essentially, the agents try to reach an agreement

11: with the constraint that it has to be approached no slower than a

12: prescribed convergence rate. Under that constraint, global

13: consensus may not be achieved and only local agreements may be

14: reached. The agents reaching a local agreement form communities

15: inside the network. In this paper, we analyze this opinion

16: dynamics model: we show that communities correspond to

17: asymptotically connected component of the network and give an

18: algebraic characterization of communities in terms of eigenvalues

19: of the matrix defining the collective dynamics. Finally, we apply

20: our opinion dynamics model to address the problem of community

21: detection in graphs. We propose a new formulation of the community

22: detection problem based on eigenvalues of normalized Laplacian

23: matrix of graphs and show that this problem can be solved using

24: our opinion dynamics model. 

25: We consider three examples of networks, and compare 

26: the communities we detect with those obtained by existing algorithms based on modularity optimization.

27: We show that our opinion dynamics model not only provides an appealing approach

28: to community detection but that it is also effective.

29: \end{abstract}

30: