InferenceEngine
Summary
Algorithm for inferring user preferences based on activity.
Details
This algorithm used probability theory to determine the most likely choice a user would make from a selection based on previous selections, without the need for collaboration or comparisons with other user activity. The demo in the screenshot shows randomly generated adverts of cars, and the estimate of the most likely car the user would choose based on their acceptance or rejection of previous ads. This could, for example, be applied to a classifieds listing website to infer the user’s preferred ads.
Screenshot
Test application. Click for a larger picture.