The embodiment of the invention provides an
abnormal number identification method and equipment. The method comprises the following steps: determining a value of each
label dimension corresponding toeach to-be-identified number through the
label dimension; according to the value of each
label dimension corresponding to each to-be-identified number, obtaining first, second and third factor variables, further, determining an estimated total value of each label dimension corresponding to each to-be-identified number; and depending on the estimated total value, determining whether each to-be-identified number is an
abnormal number or not. The label dimension comprises a
machine changing parameter, a position parameter, a card type parameter, a communication behavior parameter, a
certificate parameter, a user parameter and the like. According to the embodiment of the invention, the factor variables are constructed from the label dimensions such as
machine changing, positions, card types, communication behaviors, certificates and users, so that the abnormal numbers are identified, the label labeling of the abnormal types of the numbers is finally realized, and the problem of relativelylow
abnormal number identification accuracy caused by the fact that the existing extracted features cannot truly reflect the characteristics of the user numbers is solved.