FRAUDULENT DATA DETECTOR

摘要

An apparatus identifies suspicious data records having two or more numerical fields. A first hardware selector identifies a set of records for analysis. A second hardware selector identifies fields within identified records that are appropriate for a Benford analysis. A Benford analysis engine calculates, for each identified field, a Benford distribution for each identified field. A hardware aggregator sums a total score for each record from the set of records, where each total score comprises a summation of deviant values for each appropriate field value within each record, and where a deviant value represents a difference between a calculated Benford distribution for each field and a theoretical Benford distribution for each field. A third hardware selector selects a record from the set of records according to a highest total score.

主权项

1. An apparatus for identifying suspicious data records, said records comprising two or more numerical fields, said apparatus comprising:
a first hardware selector for identifying a set of records for analysis; a second hardware selector for identifying fields within identified records that are appropriate for a Benford analysis; a Benford analysis engine for calculating, for each identified field, a Benford distribution for said each identified field; a hardware aggregator for summing a total score for each record from the set of records, each total score comprising a summation of deviant values for each appropriate field value within said each record, a deviant value representing a difference between a calculated Benford distribution for said each field and a theoretical Benford distribution for said each field; and a third hardware selector for selecting a record from the set of records according to a highest said total score.