The scientific community organizes its relations into network patterns
, where the nodes are individuals and the links are acquaintance and c
ommon work. Scientific meetings are marketplaces of specialized knowle
dge, where new links are established by exchange of information. The f
low of knowledge is modelled by information transfer that may be calib
rated through polling the participants on the number of 'contacts' the
y establish at such meetings. We indicate estimates of cost and effici
ency of scientific meetings which may be important for helping science
and technology funding agencies in developing countries to increase t
heir community's impact in the world of science.