Knowledge Based Approach to Form of the Assessment Means Fund

The article proposes a theoretical-model approach to modeling the fund of assessment documents. This approach allows the formation of assessment documents, taking into account the wishes of the examiner. Formalization of the object domain in the form of Case Model and it Fuzzification allows you to build a template of an assessment document in the form of a quantifier-free formula of predicate logic, as well as to check the consistency of the constructed template.The article describes an algorithm for generating a set of assessment documents for a given template. This set consists of a user-specified number of variants of assessment documents of the same type in structure, but unique in content. The algorithm is based on the categorical data clustering algorithm CLOPE. The paper presents a modification of the algorithm that allows to automatically determine the required number of clusters depending on the input data, as well as to solve the problem of small clusters.