Amadeus S.a.s.

Cedric Florimond

Donald Kossmann

Laure Canis

Gustavo Alonso

Dietmar Fauser

Georgios Giannikis

Philipp Unterbrunner

Jeremy Meyer

Thibaud Andrevon

Simone Bisogni

Marion Bonnet

The present invention relates to a computer-implemented fraud method and system for managing fraud screening in response to a requested transaction. The fraud management system checks the efficiency of the fraud screening strategy and predicts the efficiency of new fraud screening strategies based on the total cost of fraud. This calculation is facilitated, because transactions are divided into groups, and the total cost of fraud may be calculated for a representative element of the group, and not for each transaction in the group. Furthermore, if the fraud screening is based on rules that apply an acceptance flow to predetermined conditions, the fraud management can choose the best acceptance flow for these predetermined conditions by minimizing the total cost of fraud. Finally, the fraud management can also flag inefficient rules thanks to an indicator related to the total cost of fraud to highlight rules whose predetermined conditions should be changed.

Fraud management system and method

Methods, systems, and computer program products for screening transactions. Transactions stored in a transaction database may provide a statistical basis for determining a probability of fraud for a pending transaction. The cost of fraud may be determined for each of a plurality of possible actions based on the probability of fraud, and the action that provides the lowest cost of fraud chosen. A cumulative cost of fraud may be determined for a set of transactions in the database. If taking the action with the lowest cost of fraud for the transaction would result in a higher cumulative cost of fraud, an action having the higher cost of fraud for the transaction may be chosen. The transaction may be added to the database, and the probability of fraud adjusted based on an amount of time since acceptance of the transaction. The statistical basis may be optimized based on a cost error.

Predictive fraud screening

A distributed data store includes a storage layer comprised of a plurality of storage nodes, each storage node configured to receive and queue table operations, and an aggregation layer comprised of aggregator nodes configured to route table operations to storage nodes in the storage layer and merge results of table operations returned by storage nodes in the storage layer. Each storage node includes in part scan threads dedicated to continuous and full scanning of segments of a table that are respectively associated with processor cores upon which such scan threads are running.

Continuous full scan data store table and distributed data store featuring predictable answer time for unpredictable workload

A method for storing and retrieving data in a storage node of a data store and storage node of a data store, storing in main-memory at least one segment of a relational table. The storage node includes at least one computational core running at least one scan thread each dedicated to the scanning of one of the at least one segment. At least one scan thread uniquely, continuously and exhaustively scans the dedicated segment of the relational table. The storage node receives and processes batches of query and update operations for the at least one segment of the relational table. The query and update operations of a batch are re-indexed at beginning of each scan by the scan thread. Then, the indexed query and update operations of a batch are independently joined to data records of said segment that match with predicates of the indexed query and update operations.

Growing community of inventors

Vallauris, France

Cedric Florimond