Research Of Product Form Preference Prediction Based On User’s Similarity

With the changing times, user’s aesthetic taste changes and the industrial design theory and technology develops. Then product form design is becoming one of the most active direction in the theory of product design. In recent years, the innovation design is referred to the unprecedented height. Companies are investing money in terms of product innovation research, of which the innovation design of product form is an essential part. Therefore, designers are dedicated to the product form design to meet the user’s aesthetic. User’s perceptual preference of product form is based on the user’s emotions and aesthetic needs. Users deal with perceptual preference through cognition- feedback process, and different users have different preferences, namely user’s perceptual preference has complexity and diversity.This article takes finding similar users as a key, to study the correlation between the user and the perceptual preference of product form, to find and predict the target user’s personalized perceptual preference of product form. The research is focus on the following several aspects.Firstly, based on the research of user’s perceptual preference of product form, we analyze the way of classifying the user groups, and discuss the meaning of user groups to the study of user’s perceptual preference of product form.Secondly, according to the characteristics of the product form, find out the calculation method of the product form similarity. Extract the product form features. Then determine the weights of the features in the perception of the product form similarity,and use vector space model to transform the presentational attribute into mathematical model, to calculate the product form similarity.Thirdly, according to the correlation between the user groups and the perceptual preference of product form, this paper proposes an algorithm to obtain similar users, in order to predict the target user’s perceptual preference of product form based on the collaborative filtering approach. Finally, realize the prediction of user’s preference of product form in the form of system.