The invention provides an assessment method and device for the risk degree of a person and electronic equipment. The method comprises the steps that target data is obtained; through the target data, person relationship networks are built; according to a target algorithm, the person relationship networks are treated to obtain characteristic vector information of each node in the person relationshipnetworks; the characteristic vector information is input to a deep convolutional neural network to determine the risk degree, of each type, of the person to be assessed. According to the assessment method for the risk degree of the person, the data resources are wide and not limited to dialog data but relates to person relationship data and consumption data, and based on the data, the multiple person relationship networks are built, so that the problem of single data resource is solved; moreover, the risk degrees, of the multiple types, of the person to be assessed can be determined, and thetypes are not limited to a certain criminal type, so that the technical problems are solved that according to existing risk degree assessment methods, the data source is single, and assessment can only be conducted on specific criminal crimes, so that the limitation is large.