Novel techniques for survey and classification studies to improve patient centered websites.

There is great interest in ascertaining patient perceptions in order to create more patient-friendly web resources. The recent proliferation of inexpensive web based data collection systems can facilitate this. Many quite sophisticated tools are commercially available. Unfortunately, researchers often recreate these capabilities in order to avoid privacy issues. This poster describes a simple architecture that allows use of a commercial system while maintaining privacy. In this example, the commercial tool supports the collection of complex categorical sorting data relating to chemotherapy systems. Hypothesis discovery techniques are used to convert the sort data into intuitive web menus.