Design And Implementation Of Personalized 3D Model Search For Home Design

With the rapid development of the Internet,information overload has become a problem that people need to face.Personalized search engines,as user-centered search engines,can not only improve information overload,but also provide users with accurate information,and have been widely studied and applied in News,E-commerce and other fields.Aiming at the problem of low search efficiency and inaccurate search in home design software,this paper implements a 3D model personalized recommendation system for home design based on the classic 3D model search system in home design software,and then integrates it into Kujiale cloud platform to realize personalized search of 3D models.Using the user’s historical operation information in the home design,this paper proposes a feature related to the context of the furniture insertion operation as the semantic feature of the 3D model in home design recommendation system.This feature can better consider the historical operation context information of the current user as a basis for sorting and recommending furniture to be inserted.In the training phase,this paper obtains the label sequence of the 3D model according to the user’s historical operation information in the excellent design solution,and based on the word2vec method,each 3D model is embedded with context-related representations of the operation.The experimental results show that the proposed personalized recommendation method can improve the efficiency of program design for Kujiale users.