Porosity is the fraction of rock volume filled by fluid. Realistic porosity values improve oil reserves evaluation, but are deeply dependent of the correct evaluation of matrix physical properties or matrix porosity parameters that can be defined as the measures accomplished by porosity tools (density and neutron porosity) in a null porosity rock. Porosity parameters are obtained by an intelligent algorithm controlling the operations of an artificial neural network, the angular minimum competitive neural network, in the interpretation of RHOb-PHIn plot. This method is able to produce the log zonation and the porosity calculated without shale correction. The presentation of this methodology is accomplished with synthetic data and evaluated using porosity logs from boreholes of Namorado Oil Field, in Campos Basin. Brasil.