Specialty goods price evaluation using social networks-based producers representation
Abstract:
Disclosed herein are system, method, and computer program product embodiments for a method of evaluating specialty goods. The method receives data representing a graph specifying relationships between producers and institutions. Based on the relationships specified in the graph and a first set of weights, the method determines a convolution value for a producer such that the convolution value inherits attributes of institutions for which the producer has a relationship. The method receives data describing a specialty good from a producer and inputs the convolution value and the data describing the specialty good into a machine learning module with a second set of weights. The first set of weights and the second set of weights are trained from known evaluations of specialty goods of the producer. From the machine learning module, the method receives a value of the specialty good.
Information query
Patent Agency Ranking
0/0