2023 年 7 巻 2 号 p. 2_31-2_38
In recent years, there has been a demand for increased diversity of ideas in product development. In the conventional design method, the brainstorming method and KJ method are used to increase the number of keywords by text and organize the relationships. However, with this method, the idea generation process takes time and the results are limited by the knowledge of the user. Therefore, in this study, to generate various ideas in a shorter time, we developed some of the main functions of a system for idea support using AI that handles large-scale text and image data. The effect of the developed system was verified from the amount of ideas generated and results of sensory evaluation experiments. The experimental results suggested that the presentation of image information is more useful in idea generation than the presentation of text information. In addition, they suggested that the idea is effective in facilitating the image.