In Social Network Analysis (SNA), k-core decomposition is used to detect hierarchical shells in networks. The application of the K-core decomposition to a network of keywords allows us to represent the conceptual structure of a research field. The objective of this work was to propose the application of k-core decomposition to show the evolution of the conceptual structure of the Altmetrics research field. The methodology was developed in several phases: data collection, keyword selection, elaboration of a keyword co-occurrence matrix, generation of a keyword network, k-core decomposition and visualization of the hierarchical structure. The result was the detection of five differentiated shells. A core shell with basic, densely interconnected concepts that formed the knowledge base of the field. An intermediate shell with mediating concepts that showed the evolution of knowledge in the field. A lateral shell with concepts that indicated the specialization of the research field. A border shell with peripheral and isolated concepts, which represented the conceptual fronts in development. In conclusion, the hierarchical decomposition of the keyword network achieved a deeper understanding of the conceptual structure of the research field