Application of fuzzy ISODATA clustering method in product identity planning

Product identity (PI) can form the continuity of enterprise products, and is of positive significance to the brand image. In this paper the fuzzy ISODATA clustering method is introduced to explore the way to establish product identity system based on the existing products of a company. An experimental study of financial equipments was conducted to evaluate the practicability of this method. First, appropriate product form features were selected and the values of each sample´s features were acquired by semantic differential method; the initial division was given by the designers´ experience, and then the result of categorization and the information of the clustering centers of each style was calculated using fuzzy ISODATA clustering method. On this basis the company´s PI system including the sub-styles can be planned. The future products can be designed with continuity according to the result. The experimental study has shown a feasible and effective result.