Botanical data retrieval system supporting discovery learning

We constructed a botanical data retrieval system applying our proposed search interface named 'Concentric Ring View' for multi-faceted metadata. This system allows users to search flexibly and intuitively by combining attributes with simple operation. The attributes used as search keys are visual and botanical features such as flower color, leaf shape, blooming season, and so on. Users can create their own dynamic knowledge hierarchies by selecting these attributes and adjusting attribute values. We considered that this system enables users not only to search for plant names but also to learn morphological features and the taxonomy of plants, and performed a usability testing. We confirmed that users could finally find the correct plant names even if each user selected different attributes for searching, and users could also find correct plants by grasping visual features even if the images shown focused on flower and leaf. In discovery learning, users could learn plant features and botanical properties by finding out the common properties from the relationships between attributes and attribute values, and attribute values and retrieved results.