The second part of this review deals with the methods of inverse kinetic problem (IKP) solving based on generalized descriptions of the process. The application of these methods is considered an alternative (for discrimination) approach to IKP solving. It is shown that the base of this approach is the methodology of complementarity. Different types of generalized descriptions, their merits and their shortcomings are discussed.As noted in the first part of this review, the unambiguous description methods reduce the ambiguity of IKP solutions through discrimination by applying additional information. This information is regarded as external with respect to the data of a particular kinetic experiment, since it is only derived within certain assumptions of specific experimental conditions (on the distribution of random quantities, process nature, IKP solution properties, etc.). The applicability of these assumptions stems from general theoretical concepts rather than from the analysis of the kinetic curves described. In contrast with the discrimination methodology, the complementarity methodology implies the fullest possible use of internal information (i.e. that obtained in the particular experiment). Besides the difference in the origin of the information used, the complementarity methodology is characterized by the description of a process synthesized in the course of the IKP solution rather than presented by the a priori set "rigid" system of formal models. The methodological basis for generalized descriptions, their structure and their relationship with the IKP solution are considered in the following section.