With the growing demand for immersive and high degree- of-freedom content, traditional two-dimensional (2-D) scene content diction algorithms exhibit significant limitations in three-dimensional (3-D) Holographic communication environments. Current recommendation mechanisms primarily focus on content tag matching while lacking coupling optimization between spatial interaction characteristics and functional decision-making in holographic scenarios. To address this issue, This paper proposes a content-function coupling recommendation mechanism for holographic communication, innovatively integrating 3-D con-tent characteristics with immersive interaction requirements. First, we employ a long- and short-term memory (LSTM) model to capture the spatio-temporal behavioral sequences of the users, combined with collaborative filtering (CF) algorithms to extract global static preferences. Second, a lightweight multilayer perceptron (MLP)-based functional decision model is designed to determine transmission parameters according to the terminal network status and computational capabilities, proactively selecting combinations of parameters to reduce latency and minimize resource waste. The experimental results demonstrate that, compared to traditional recommendation systems, the proposed mechanism achieves significant improvements in both the transmission stuttering rate (60%) and the transmission speed variation latency (70%).