Final qualifying work by Ermolaev A.N. "Development of methods and algorithms for social network analysis using graph-theoretical approaches" is devoted to the construction and analysis of graph models of communication in the communities of users of the social network Vkontakte. The content of the work is fully consistent with the theme stated in the title. The structure of the work is chosen in accordance with the sequence of tasks formulation, including theoretical, practical (program) and experimental parts.  In recent years, the study and modeling of social networks has become very popular in various fields of science, including mathematics using the apparatus of graph theory and game theory. Social networks have become firmly established in our life in the last two decades, and during this time there was a huge number of scientific works devoted to these studies, including those that were used by the author in the performance of this qualification work. Therefore, the relevance of the work is beyond doubt. At the beginning of the work is an introduction, the formulation of the tasks and a brief review of the literature on the subject, rather superficial. The first chapter presents the basic terms and formulations of graph theory used in this paper, including a description of the structure of graphs modeling communities in the social network in this paper and the value of the metrics calculated in the work for the graphs (density, clustering coefficient and the assortability coefficient). The second chapter contains a technical description of the developed software tools for automated collection of information about the structure of the studied graphs, their visualization and finding them for the values of metrics. The third chapter describes the experiments conducted on real data of social network subnetworks Vkontakte. At the end of the work some conclusions are made concerning the results of experiments. There are a number of questions and comments to the work. 1. Not enough justification for why the three metrics (density, clustering coefficient and the coefficient of assortativity) were selected for the study. 2. Description of the concept of density, unlike other metrics, is given without reference to the source. Whether in this case the notion of personal contribution of the author in the development of science? 3. The authority of a number of sources listed in the list of references is questionable. In particular, these are electronic resources 9, 10, used as a basic source of general concepts, history, metrics and methods of analysis of social networks. Since the final qualifying work claims to some scientific content, then as sources should be used more serious scientific literature than anonymous articles on the Internet. 4. The results of experimental studies are poorly interpreted. For example, in reality, the distributions of metric values can be denoted for the studied communities (Fig. 6-9)? How to interpret the resulting extreme values in the Table. 1, including negative assorted? 5. A significant part of the text of the work (25%) is unclear why the above listing of the program – 9 pages of uninformative code without a single comment. What is new and useful according to the assumption of the author, the reader can find for himself in this text? In view of all the above, the overall impression of the work is as follows. The author has successfully completed the practical tasks set in the work, however, the theoretical material, as well as the results are presented superficially. I think that this final qualifying work deserves the evaluation of "Good".