Social and spatial interaction models

The flow of news produced by media or individuals located in a given place about conflicts located in another place can be explained in the majority of cases by variables related to continuous measure of proximity, and network structure. Modelling these effects is crucial in order to filter the banal part of flows - related to general rules - and the specific effect of singular conflicts or event - revealed by positive and negative residuals. The deliverable will consist in a technical report on the use of spatial and social interaction model for the filtering of conflicts in news of flows. It will be associated to a set of programs and a selection of datasets for examples, that could take the form of a package in an open source language like R.