The Bacground of the Development of CAKE

User experience is one of the most important issues in design. Lots of methods have been proposed for experience design, among which, Kansei Engineering is frequently employed for helping designers understand usersˇ¦ preference and applying it to products. With the help of advanced computer technologies, several computerized Kansei Engineering systems have been developed. However, due to the arbitrary format of data representation schemes and the absence of integration mechanisms, it is not easy to merge or to compare the results between different studies. For this problem, a comprehensive CAKE (Computer-Aided Kansei Engineering) system with extensibility is developed by using the XML (Extensible Markup Language) technology.

Figure 1: The overall concept of the CAKE System

The Architecture and Design of CAKE System

There are two major components of this system: tools for survey and for data analysis. The survey tool facilitates researchers to design an interactive experiment and use it to collecting usersˇ¦ preference data efficiently and effectively. By visualizing the results of frequently used statistics, the analysis tool assists researchers to examine the data and accumulate the findings into an XML file. In order to improve the efficiency and accuracy of data collection, there are six types of interactive survey tools developed. Among them, there are two tools for investigating large set of stimuli. By conducting the experiments to compare with the traditional approaches, the improvement of efficiency and reliability of these computer-based methods was investigated. This system has been applied to several studies with products, such as automobiles, sofas, and kettles.

Figure 2: The Architecture and Design of CAKE System

  1. Computer Aided Kansei Survey

    In this module, a computer application is developed to help researchers use XML to construct an interactive kansei survey. After CAKES is launched, users can select any function from the wizard window. Currently, the tools for Aiding Kansei Engineering Survey have been implemented, and detail descriptions are given bellow.

    Figure 3: The wizard of the CAKE System

    After this tool is activated from the wizard window, an ease-to-understand animation will demonstrate the process in creating a new Kansei Survey through three steps: 1) Import and define the stimulus, 2) Input pairs of adjectives, 3) Select types of participantsˇ¦ information needs to be collected.

    Figure 4: A example of designing the survey with CAKE
  2. Interactive Survey Tools:
    By using the internet based interactive Kansei Survey, the quantity and quality of data collection can be improved by the usage of several computer technologies. For example, the sequence of the semantic difference scales can be set randomly from subject to subject. Together with the interactive rating mechanism, better survey results could be obtained.
    Figure 5: An example of rating with the survey tool.
    Figure 6: The comparison of different survey tools.

  3. Data Visualization:

    When the data collection is completed, a researcher can simply parse the scores into several formats for running analysis with statistic softwares, such as SPSS or MDPref, a computer program frequently used for analyzing multidimensional scaling data. After the researcher gets the results, he/she could use the Data Visualization function to render the quantitative results. By utilizing computer interactivity and the graphical representation advantage [Tufte, 2001], it was found that this tool could help the researchers draw the findings more quickly and accurately. At the same time, the statistic data and the findings a researcher concludes from the visualization will be automatically accumulated into the XML file. In the future, when a user accesses the study from the CAKEˇ¦s web system, he/she can get all of the rich information for designing the new product or conducting a follow-up research.

    Figure 7: An example of analyzing the survey data with the visualization function.