Apply fuzzy case-based reasoning to knowledge acquisition of product style

Extracting product style from morphological features is an important method for style knowledge acquisition. And the repeated use of the similarity of design patterns in shape, color and other factors of product at design stage is the main factor of the formation of the product style, which makes the case-based reasoning method suitable on style extraction research. Due to the uncertainties in representation, attribute description, and similarity measures of knowledge in product design issue, a fuzzy case-based reasoning (FCBR) was developed in product style extraction by using linguistic variables. A product is encoded by a vector consisting of many attributes, and product morphology database was established subsequently. Fuzzy K-Nearest Neighbors (FKNN) was introduced to define the style similarity between two products. The model for product style extraction was generated employing the proposed FCBR system, and the results were normalized by Fuzzy Sets. Experimental results show the effectiveness of the FCBR model when comparing it with other approaches in product form style extraction.