The growing recognition of attributed networks as a crucial element in the field of data science stems from the growing abundance of this type of data, particularly within network contexts [1, 2, 3, 4]. Our contribution presents a novel approach to clustering attributed graphs. It proposes an objective function that uses regularized fuzzy clustering to enhance the quality of embeddings while ensuring effective clustering of nodes within graphs. Therefore, instead of treating feature information X and node topology W as distinct entities, we propose to base ourselves on an objective function that integrates embedding and fuzzy clustering. Using the characteristics of a low-rank subspace and fuzzy clustering, our method aims to capture the intricate connections between X andW, thus improving the robustness of clustering. Experiments are carried out on benchmark-attributed networks of different sizes to assess how well our algorithm performs compared to leading clustering methods designed for the same task.