Privacy has been recognized as a critical topic in the Internet for a long time, and technical developments in this area are ongoing. With the proliferation of various network sites and the unification of them, it has become more evident that the problem of privacy is not bounded by the perimeters of a specific system. Nowadays, many network services rapidly drive toward unification. In this paper, we mainly consider unified personal information. We propose an evaluation method of privacy sensitivity-level. The information sensitivity level (S(L)) is evaluated with four main factors; the degree of conjunction (D(C)), the principle of identity (I(A)), the principle of privacy (P(A)), and the value of analogism (A(A)), S(L) := Sigma(D(C), J(A), P(A), A(A)). If we classify the information based on the information sensitivity level (SO, it can show the objective validity. Finally, we outline new research directions for currently existing methods for privacy-preserving data analysis in the combined social network environment