The invention relates to a method of predicting personality based on data generated from multiple data sources and determining a recommended item based on the predicted personality. A platform server receives, from a plurality of source devices, personality data associated with a user. Using the received personality data, the platform server generates a set of diagnostic features. At least one model is applied to the set of generated diagnostic features to generate at least one personality measurement. The personality measurements are used to generate a personal profile. The personal profile is matched to at least one recommended item, and the recommended item is transmitted to the user.
展开▼