Co-authorship networks reveal who collaborates but not the intellectual content of research. We integrate co-authorship data from the Italian Computer Science community (INFO-01) with OpenAlex topic annotations to capture both structural and thematic dimensions of collaboration. Using 2024 data, we map how subfields interact internally and with external domains, highlighting the central bridging role of Artificial Intelligence. We also compare graph-based, topic-based, and combined embeddings, showing how semantic and structural similarities complement each other. This integrated approach provides a richer understanding of intra-disciplinarity and offers tools for identifying potential collaborations and emerging research areas.