: The problem of adjustment of modern intelligence enhancement methods and automated data analysis methods to the problems that are still handled manually is fairly topical. For the solution of such problems, this study suggests a new DT representation which uses approximated to the NI knowledge structuring. The structuring is implemented by the authors’ question-answer binary tree. This is a new DT with only most optimal decisions for all known situations excluding non-efficient cases. A set of ‘the most effective’ solutions are leaves of the tree. This new approach can be applied in intelligent decision support systems (IDSS) which enhance the natural intelligence of the scientist in the exploratory research. This tree was tested on the problem of selecting ‘the most suitable’ optimization method out of all known ones. First, detailed material on the main optimization methods was selected. The material was processed by new rules of deriving tree elements. The resulted tree has 127 nodes, 64 leaves are optimization methods (solution options). 63 intermediary nodes form a unique path from root to leaf, showing the progress to the most suitable method. Also, an IDSS was implemented in C#. The paper dwells on all stages of the DT construction with detailed illustrations, including video. The suggested DT allowed: simplify knowledge base designing; reduce system designing time; simplify decision search algorithm in the knowledge base; refer to the expert in case of contributing one’s own developed knowledge to the subject field in the tree; obtain a new way of meta-knowledge representation.