Adaptive Probabilistic Semantic System and Method

摘要

An adaptive probabilistic semantic computer-implemented system and method links semantic chains comprising subject, predicate, and object triples and an associated probability with behavioral-based chains comprising a system user, predicate, and object triples, and infers user preferences based on the linked chains. Recommendations are generated for delivery to system users based on the preference inferences. Explanations may be provided to recommendation recipients as to why they received recommendations, and the explanations may convey a sense of confidence in accordance with probabilistic semantic associations.

主权项

1. A computer-implemented method:
accessing automatically a first behavioral chain comprising a representation of a first user of a computer-implemented system, a first predicate, and a first object of the first predicate; accessing automatically a first semantic chain comprising a first subject, a second predicate, a second object of the predicate, and a first semantic chain weighting associated with the first semantic chain, wherein the first subject corresponds to the first object of the first predicate; linking automatically the first semantic chain and the first behavioral chain; inferring automatically a preference of the first user of the computer-implemented system based, at least in part, on the linked first semantic chain and first behavioral chain and the first semantic chain weighting; and selecting automatically a recommendation for delivery to the first user of the computer-implemented system, wherein the recommendation is selected based, at least in part, on the inferred preference.