감성공학을 적용한 자동차 운전석 디자인 프로세스 모델에 대한 사용성 평가

This paper describes a neural network based website Kansei usability evaluationprototype system. This system simultaneously operates with the Microsoft^TM Internet Explorer, and the results are learned using a neural network. In this paper, firstly, we derived the Kansei adjectives and the effective website usability factors, and they were matched by Correspondence Analysis. Then, highly corresponded adjectives were implemented on the system for the Kansei evaluation. Also, the evaluation efficiency of the suggested system was investigated using SSE (Sum of Squared Error) and estimated user`s satisfaction. Finally, the results showed the efficiency of developed algorithm and system for the website evaluation.