A Visualization Method for Helping Children Assess the Risk of Websites

Since children cannot live in a safe cyberspace environment for ever, setting up filtering or using educational materials are commonly used. A more effective approach to helping children assess website risk is to provide an environment in which children can better understand a website's features and determine the risks of accessing the website for themselves. We have constructed a prototype visualization system for helping children understand website features and use them to identify high-risk websites. We applied a graphical search interface named 'Concentric Ring View' which we proposed to support flexible retrieval for multi-attribute metadata. It was tested using actual and dummy websites featuring five attributes: action, color, atmosphere, number of images, and number of links. The risk level of each actual website was estimated by the features of the portal sites through which it was accessed. A dummy website was used when the actual one was deemed to be too risky. The testing revealed several distinguishing characteristics of high-risk websites. Adding other search attributes such as structure and text appearance should make it possible to characterize risky websites more completely.