Goal. Application of fundamentally new approaches to solving the problem of structural syn-thesis of ill-defined objects is, on the one hand, a challenge of time that requires reducing the time for development and obtaining high-quality technical solutions, and, on the other hand, the result of research that allows the use of invariant trial and error algorithms for structural synthesis. Solving this problem by the proposed method requires the creation of the set of data and knowledge, the search of which will give the desired result. Formation of this set and control of its correct and actual state is a serious task requiring the development of new approaches, formalization of processes and development of algorithms imposed by these approaches. Methods. The stages of formation of the pointed set are: collection of data on existing technical solutions, their formal description, gene-ralization of these descriptions with the formation of a certain structure, which allows for identifying each existing object unambiguously and creating a new one either by a combination of existing knowledge or by upgrading the known technical solutions. It is proposed to use a classifier as a structure that summarizes the data of a specific area of knowledge. The peculiarity of this classifi-er is that two methods of development are implemented here: hierarchical and faceted. Creation of such a classifier and its correct state is a problem that allows to solve the problem of formation of information support for structural synthesis systems of ill-defined objects of medium and high complexity. Results. The process of creating such a classifier deals with a number of problems, the identification of which is the first step towards the development of high-quality information support. These problems arise at each stage of the domain analysis (analysis of existing versions of objects; generalization of knowledge about the considered objects; addition (extension) of the classifier). The reasons for their occurrence may be violation of the principle of sufficiency in analysis and de-composition, violation of the hierarchy of the structure when generalizing in the process of building a domain model, violation of the signs of mandatory and optional when synthesizing and expand-ing the model, which is based on the classifier. Conclusion. Identifying these problems and studying the process of avoiding and overcoming them is an important task, the solution of which will allow you to for-malize the process of creating a classifier as an information model of a structural synthesis system.