Reducing false positives with transaction behavior forecasting
AssigneeBrighterion, Inc.
InventorAkli Adjaoute
Life of this patent
Estimating the term…
Abstract
An artificial intelligence fraud management system comprises a real-time analytics process for analyzing the behavior of a user from the transaction events they generate over a network. An initial population of smart agent profiles is stored in a computer file system and more smart agent profiles are added as required as transaction data is input. Transactions in particular merchant category codes (MCC) are likely to be followed by predictable related transactions. A forecast of those likely future transactions is calculated and used to desensitize corresponding smart agent profile datapoints. Fewer false positives are produced and overall fraud management performance is improved.
Claims20
A computer-implemented method for real-time analysis of a series of transaction events reported over a financial network and corresponding to a plurality of transacting entities, comprising:
- automatically receiving, at one or more processors, a series of transaction records corresponding to the series of transaction events and including real-time transaction data;
- automatically timestamping, via the one or more processors, the real-time transaction data of the series of transaction records;
- automatically locating, via the one or more processors accessing respective profile blocks with meta-data headers, a plurality of vectors respectively corresponding to the plurality of transacting entities, each of the plurality of vectors reflecting historical data and pointing to a plurality of datapoints that:
- (A) are represented individually in or are derived from combinations of data types comprising the real-time transaction data, and (B) share a common time interval;
- automatically reassigning, via the one or more processors, at least one of the plurality of vectors having an expired common time interval to a new common time interval;
- automatically analyzing, via the one or more processors, all vectors of the plurality of vectors that correspond to an instant transacting entity in connection with scoring an instant transaction record of the instant transacting entity;
- based at least in part on the analysis, automatically adjusting, via the one or more processors, a transaction risk score;
- and
- automatically outputting, via the one or more processors, a fraud score based at least in part on the transaction risk score.
A server for real-time analysis of a series of transaction events reported over a financial network and corresponding to a plurality of transacting entities, the server comprising:
- one or more processors;
- non-transitory computer-readable storage media having computer-executable instructions stored thereon, wherein when executed by the one or more processors the computer-readable instructions cause the one or more processors to—
- automatically receive a series of transaction records corresponding to the series of transaction events and including real-time transaction data;
- automatically timestamp the real-time transaction data of the series of transaction records;
- automatically locate, via accessing respective profile blocks with meta-data headers, a plurality of vectors respectively corresponding to the plurality of transacting entities, each of the plurality of vectors reflecting historical data and pointing to a plurality of datapoints that:
- (A) are represented individually in or are derived from combinations of data types comprising the real-time transaction data, and (B) share a common time interval;
- automatically reassign at least one of the plurality of vectors having an expired common time interval to a new common time interval;
- automatically analyze all vectors of the plurality of vectors that correspond to an instant transacting entity in connection with scoring an instant transaction record of the instant transacting entity;
- based at least in part on the analysis, automatically adjust a transaction risk score;
- and
- automatically output a fraud score based at least in part on the transaction risk score.
Dependencies are inferred from the claim text.