In the modern world in various spheres of activity, the number of problems that need multi-criteria decision-making (MCDM) is constantly increasing. Researchers and experts offer a number of approaches to MCDM process in such tasks; in particular, most of them are based on expert methods. However, in practice, these methods require significant intellectual effort of experts and organizational and technical workload during the expert survey, and also usually take a long time. At the same time, it is not always possible for experts to use certain characteristics of alternatives, which also carries the risk of making decisions based on unfounded expert opinions. Therefore, such methods and tools should be clear and informative and at the same time easy to use to ensure the efficiency and effectiveness of their use. We offer a graph-based approach to expert decision-making and information visualization processes that meets these requirements and can significantly improve the efficiency of decision-making in multi-criteria selection tasks.