Patent attributes
A real-time recommendation system includes a non-transitory, computer-readable storage medium having encoded thereon instructions that a processor executes to receive data from data sources that includes one or more data records. The processor then identifies features in data of a data record, extracts and buffers the features, and applies the features to a feature detect mechanism. The feature detect mechanism includes feature detect algorithm and more recommender algorithms that are encapsulated in a feature detection object. The feature detection object may be a Strategy/Composite feature detection object that includes Strategy design patterns. The processor further executes to reject a feature in response to failures of the feature to match at least one of the Strategy design patterns, and in real time, select and install features that match one or more of the Strategy design patterns into one or more recommender algorithms.
