For a collection of agents, information systems, or knowledge components, the level of interoperability and the preferred interoperability structure can be assessed by analyzing the tendency to establish relationships between them. As one possible implementation approach, the concept of structural consistency is proposed. Structural interoperability is based on relationship analysis using a consistency criterion that ensures an optimal level of coherence for the analyzed set of elements. The minimal deviation from an identified consistent structure corresponds to a preferred composition in which interoperability can be achieved with the highest degree of motivation. As a result, a set of potentially interacting elements is partitioned into groups of elements that are mutually motivated to interact. This paper proposes and substantiates an algorithm for determining the closest consistent structure for an arbitrary set of elements, which makes it possible to justify the choice of an appropriate interoperability structure. Examples illustrating the application of the proposed algorithm are provided.