Abstract: A data processing system is configured to identify treatment responsive to a health risk determined from feature data provided by one or more networked data sources. A classification engine generates a feature vector based on a natural language processing (NLP) of input data representing words provided by a user. Features of the feature vector represent health risk factors. Machine learning logic classifies the features to generate a classification metric indicating whether the features are indicative of health risks or not indicative of health risks. A prediction value is generated indicating a likelihood of each health risk factor for the patient. The patient can be diagnosed with a health condition or disease based on the identified health risks.