The dialectical approach to the knowledge representation in general form consists in the possibility of synthesizing new knowledge based on the explication of meaningful contradictions in the antinomic poles of dichotomies. The scientific novelty of the research lies in the fact that the dialectical approach is combined with new information technologies. This is how the “bridge” of reasoning is thrown from the concept of “knowledge management” to machine learning based on large, expertly labeled data. This is the basis of the synthesis carried out in the expert-metric approach, which is a special case of the dialectical approach. Now, with the development of the Internet, big data, there is a technological opportunity to use the creative potential of dialectics more fully. The basis of the synthesis depth is the representation of the management decision-making process by a set of decision trees. In this case, the metric approach is the use of a small number of decision trees, often presented explicitly, and the expert approach is the use of a very large number of trees that exist only in the head of an expert in an implicit, non-reflected form.