On Combining Partial and Incompatible Information in E-negotiation and E-arbitration

Abstract

The complexity of the problems to be addressed in an e-democracy framework and the variety of involved stakeholders, with different backgrounds, views and access to information sources, lead to consider the case in which enegotiation should be performed among subjects who have partial, sometimes incompatible, information and can hardly be gathered to discuss issues altogether, under the supervision of a facilitator. We propose a statistical method which addresses the issue of partial and incompatible information, merging it and then using it to get a final decision, possibly in an automatic way, through the processes of e-negotiation and e-arbitration.

Cooperation via sharing of probabilistic information

“…The participant then needs a methodology, which effectively solves the problem how to combine (merge) fragmental information pieces provided by its neighbours. This problem has been addressed repeatedly, for instance, in connection with probabilistic expert systems (Cowell et al, 2003), knowledge elicitation, (O'Hagan et al, 2006), cooperation of participants (Andrýsek et al, 2007;Kárný et al, 2007); etc. However, none of the existing solutions seems to be complete and automatic enough to be applied in the cooperation task concerned.…”

On a partially specified Bayesian nonparametric model for earthquakes occurrences

“…Another popular way to relax the parametric assumption is provided by the Bayesian nonparametric approach, which has seen a lot of contributions and applications since the publication of the paper by Ferguson (1973) on the Dirichlet process. Another, very recent proposal by Andrysek et al (2008) considers the case in which different sources of information lead to specify some partial and incompatible marginal distributions about components of random vectors, that is, all of them are specified on proper, in general different, subsets of the random variables and there exists no joint distribution having all of them as marginals.…”

Processing of Information Microgranules within an Individual’s Society

“…The feature is not bounded to a given scale. We may state similarity or fuzzy relations in general between observations of a phenomenon, yet also on models describing the phenomenon, for instance in terms of behavior distribution [26] or collaborative fuzzy clustering [1,27], etc. The cooperation request, and social computing as its implementation, comes from the definition of information granularity.…”