On the Reconstruction of a Boundary Value Problem from Incomplete Nodal Data