With the continuous improvement of consumption level, Youngster has become an important group in home consumption, and their expectations for interior design are becoming personalized, rational, and efficient. Traditional design methods mainly rely on the personal experience of designers, which has many shortcomings and cannot meet the needs of people for interior design in the new era. To address these issues, this study proposes a hybrid recommendation algorithm, HRS-ID, which integrates user preferences, spatial features, and objective constraints. The core idea is to use a latent semantic model (LFM) to analyze users’ potential needs, encode spatial relationships through graph convolutional networks (GCN), and use attention mechanisms to weight and integrate the evaluation results of user and spatial dimensions to achieve accurate recommendation of design elements. In the experiment, the IKEA dataset and actual demand data from 200 homeowners were used to compare HRS-ID with other models. The results showed that HRS-ID performed well in multiple indicators: a user satisfaction score of 4.6 (out of 5), a style consistency of 92% in design rationality, a functional satisfaction rate of 95%, and a calculation efficiency of only 18 milliseconds per round of recommendation.
Yu Wang (Thu,) studied this question.