This deliverable presents the initial results of the 6G-GOALS project in the context of semantic data acquisition, representation, and compression methodologies. It marks a significant step toward realizing the vision of goal-oriented and AI-native communication in future semantic 6G networks. In the first part, the focus is placed on developing learning-based and topology-aware approaches for data representation, leveraging advanced techniques such as graph neural networks, knowledge graphs, and diffusion models. The deliverable further introduces innovative semantic compression schemes, including opportunistic and reinforcement learning-driven methods tailored to communication environments with stringent latency and efficiency constraints. In the second part, semantic data acquisition strategies are also explored, targeting practical scenarios such as multimodal sensing and opportunistic sampling. These results serve as foundational components for a holistic 6G semantic communication system where information is processed, transmitted, and interpreted according to its relevance to predefined goals.