Heuristic money laundering detection engine
AssigneeState Farm Mutual Automobile Insurance Company
InventorsElizabeth Flowers, Puneit Dua, Eric Balota, Shanna L. Phillips
Life of this patent
Estimating the term…
Abstract
A heuristic money laundering detection engine includes capabilities to collect an unstructured data set, such as a transaction record, and detect indications of money laundering activity. By detecting money laundering activity and feeding back indications of money laundering transactions, the heuristic algorithm may continue to learn and improve detection accuracy. Such indications may include correlations to sets of transaction activity among a number of financial accounts and past indications of money laundering activity. Indications of money laundering may allow generation of audit reports for reporting to regulatory authorities.
Claims20
A computer-implemented method, executed with a computer processor, that generates an indication of money laundering activity, comprising:
- retrieving, with the computer processor, an unstructured transaction set stored in a first memory, the unstructured transaction set including:
- aggregated transaction data for a plurality of users, and
- at least one indication of prior money laundering activity;
- receiving, with the computer processor, data indicative of a plurality of financial transactions from a network interface device;
- accessing, with the computer processor, a heuristic algorithm stored in a second memory;
- executing the heuristic algorithm, with the computer processor, to generate a predicted indication of money laundering activity using the unstructured transaction set and the plurality of financial transactions;
- generating, with the computer processor, a compliance report using the predicted indication of money laundering activity;
- and
- training, with the computer processor, the heuristic algorithm using machine learning, wherein the training updates the heuristic algorithm based at least in part on:
- the predicted indication of money laundering activity, and
- a correlation between the predicted indication of money laundering activity and at least one transaction identified as being associated with the prior money laundering activity in the unstructured transaction set.
A computer system configured to generate an indication of money laundering activity, the computer system comprising:
- one or more processors;
- and
- a memory storing machine-readable instructions that, when executed by the one or more processors, cause the computer system to:
- retrieve an unstructured transaction set stored in a first memory, the unstructured transaction set including:
- aggregated transaction data for a plurality of users, and
- at least one indication of prior money laundering activity;
- receive data indicative of a plurality of financial transactions from a network interface device;
- access a heuristic algorithm stored in a second memory;
- execute the heuristic algorithm to generate a predicted indication of money laundering activity using the unstructured transaction set and the plurality of financial transactions;
- generate a compliance report using the predicted indications of money laundering activity;
- and
- train the heuristic algorithm using machine learning, wherein the training updates the heuristic algorithm based at least in part on:
- the predicted indication of money laundering activity, and
- a correlation between the predicted indication of money laundering activity and at least one transaction identified as being associated with the prior money laundering activity in the unstructured transaction set.
A non-transitory computer readable medium, comprising computer readable instructions that, when executed by a computer processor, cause the computer processor to:
- retrieve, from a first memory, an unstructured transaction set comprising:
- aggregated transaction data for a plurality of users, and
- at least one indication of prior money laundering activity;
- receive data indicative of a plurality of financial transactions from a network interface device;
- access a heuristic algorithm stored in a second memory;
- execute the heuristic algorithm to generate a predicted indication of money laundering activity using the unstructured transaction set and the plurality of financial transactions;
- generate a compliance report using the predicted indication of money laundering activity;
- and
- train the heuristic algorithm using the machine learning, wherein the training updates the heuristic algorithm based at least in part on:
- the predicted indication of money laundering activity, and
- a correlation between the predicted indication of money laundering activity and at least one transaction identified as being associated with the prior money laundering activity in the unstructured transaction set.
Dependencies are inferred from the claim text.