The present invention relates to a method for predicting a human personality based on data related to a location having a specific data type, capable of providing a technology of extracting the personality of a person using a Deep Belief Network. Disclosed is a personality estimating method of estimating personality using an artificial neural network model learned using learning data including input layer-information on stay time in a plurality of locations and output layer-information representing a personality value. The input layer-information includes N*M data E_i, j (i is a natural number of 1 to N, j is a natural number 1 to m) including information on N locations and M time indexes. The value of the E_i, j for arbitrary numbers p and q (p is a natural number of 1 to N, and q is a natural number of 1 to M) is a binary number determined depending on whether a specific person is placed in a location of p at a time point of q.;COPYRIGHT KIPO 2016
展开▼