Research on recommended model of personalised interior environment design

Personalised recommendation systems are a common feature of e-commerce, streaming services, and other fields. However, they are not common in the field of interior environment design. Existing international research tends to prioritise deep learning–based recommendation algorithms. These methods rely heavily on large-scale data and complex models to improve the accuracy of recommendations. However, they often face significant limitations. Therefore, this paper proposes a personalised recommendation model for interior environment design based on clustering collaborative filtering algorithm and user portrait. The model improves the traditional collaborative filtering recommendation algorithm. Based on the established user profile and improved collaborative filtering recommendation algorithm, the constructed model is verified through performance analysis and practical application test. The experimental results show that when the number of neighbours is 20, the MAE of the model is 0.7035, which has better personalised recommendation performance. In addition, in the personalised recommendation application of interior environment design, the click-through rate of the model is increased by 0.115, and the purchase conversion rate of users is increased by about 0.270, which further proves the effectiveness of the model. The results show that this model can provide more accurate, diversified, and personalised indoor environment design solutions to meet the individual needs of users.