Intermediate nodes which are closer to the user, on the basis of information known to them but not to the source node, such as demographic data on household incomes and interests, can then modify the distribution rules to target users who may be interested in the data.