Method and system for streamlined auditing
Methods, systems, and program products for streamlined auditing that receive an input audit request via the data interface; source entity type data (ETD) from one or more databases; prepare the ETD for input into an entity clustering module; match the ETD via the entity clustering module to locate linkages within the ETD and discover relationships amongst one or more entities identified within the ETD; cluster datapoints in the ETD that refer to the same real-world entities; analyze the ETD relationships via an entity intelligence module to identify and segment targeted entities, from the one or more entities, that are applicable to the audit request; generate inclusion lists of those targeted entities that are determined to fulfill the audit request; finalize the inclusion lists of targeted entities that fulfill the audit request to generate streamlined audit results; and output the streamlined audit results to an end user.
1. A computing apparatus comprising:
one or more processors;
a memory;
a data interface; and
one or more logic element modules stored in the memory and configured to be executed by the one or more processors to perform:
receiving at a central processing unit an input audit request via the data interface;
based on the input audit request sourcing entity type data (ETD) from one or more databases via a sourcing program module of the one or more logic element modules, the ETD identifying one or more entities;
mining the sourced ETD via an entity clustering engine to identify data point linkages amongst the ETD;
generating a graph of data via the entity clustering engine using ETD data points and the mined data point linkages, whereby vertices of the graph comprise the ETD data points and N-dimensional edges of the graph comprise the mined data point linkages;
clustering the ETD data points via the entity clustering engine using the mined data point linkages via a parallelized hardware component that updates a sparse representation of the ETD graph in-place, wherein the clustering comprises update steps, whereby at each update step, the sparse representation is modified in place to have nodes, clusters of original data points, and aggregated edges to preserve structure of the graph, the entity clustering engine comprising an interactive processor that receives feedback from an end user to the ETD data points comprising rules and constraints for merging, splitting and re-clustering the ETD data points and updates and revises clusters of ETD data points;
generating initial inclusion lists using the clustered ETD data points of those targeted entities that are determined to fulfill the audit request;
finalizing the initial inclusion lists of targeted entities that fulfill the audit request to generate streamlined audit results; and
outputting the streamlined audit results to the end user to fulfill the input audit request.
2. The computing apparatus of claim 1 , wherein the ETD is retrieved from one or more public databases, private databases, and/or both public and private databases.
3. The computing apparatus of claim 1 , wherein the mining of ETD is performed for N-dimensional edges, whereby each dimension of the N-dimensional edges corresponds to a different type attribute of the ETD.
4. The computing apparatus of claim 1 , further including normalizing the ETD prior to identifying linkages amongst the ETD.
5. The computing apparatus of claim 1 , wherein the linkages are mined in the entity clustering engine using a blocking strategy and edge-scoring functions to assign a weight to said linkages.
6. The computing apparatus of claim 5 , wherein the edge-scoring functions are learned from data using Supervised Machine Learning techniques.
7. The computing apparatus of claim 6 , wherein the clustering resolves entities via Unsupervised Machine Learning techniques.
8. The computing apparatus of claim 7 , wherein the entity clustering engine is a hybrid Supervised and Unsupervised Machine Learning clustering system that builds relationships amongst one or more entities identified within the ETD.
9. The computing apparatus of claim 7 , wherein the Supervised Machine Learning training approaches generate pairwise N-dimensional similarity function regressors that enable clustering.
10. The computing apparatus of claim 1 , further including using cloud computing during one or more steps of the streamlined auditing engine.
11. The computing apparatus of claim 1 , wherein in preparing the sourced ETD includes one or more of translating, validating, verifying and/or scrubbing sourced ETD data files to enhance and prevent any erroneous data from entering the entity clustering engine.
12. The computing apparatus of claim 1 , wherein the one or more entities is selected from the group consisting of businesses, organizations, associations, enterprises, institution, establishment, governmental bodies, operation, structure, system, third-parties, and internal analytics.
13. The computing apparatus of claim 1 , wherein the one or more entities is a person or a biological being.
14. The computing apparatus of claim 1 , wherein the ETD is selected from the group consisting of entity hierarchical structures, entity legal and organizational configurations, entity operational status, entity legal status/information, trade or purpose of entity, Doing Business As (DBA) names, addresses, principals, industry or field within which entity operates, tax information, articles of incorporation, Security Exchange Commission (SEC) filings, insurance information, and combinations thereof.
15. The computing apparatus of claim 1 , wherein the ETD is associated with a person or a biological being.
16. The computing apparatus of claim 1 , wherein the end user may be an auditor, a state or federal government agency, governmental regulators, a business, a third-party entity, internal analytic departments, or any entity that may request an audit or auditing information.
17. The computing apparatus of claim 1 , further including receiving input end user feedback into the entity clustering engine to mine ETD, whereby the end user feedback is used to identify the data point linkages amongst the ETD.
18. The computing apparatus of claim 1 , wherein the one or more entities identified within the ETD comprise a list of potential auditees that may be non-compliant, in non-compliance, and/or be applicable to the audit request.
19. One or more tangible, non-transitory computer-readable storage mediums having stored thereon executable instructions for providing a streamlined auditing engine to:
receive an input audit request via the data interface;
source entity type data (ETD) from one or more databases;
mine ETD via an entity clustering engine to identify data point linkages amongst the ETD;
generate a graph of data using ETD data points and the mined linkages, where vertices of the graph comprise the ETD data points and N-dimensional edges of the graph comprise the mined data point linkages;
cluster the ETD data points using the mined data point linkages via a parallelized hardware component that updates a sparse representation of the ETD graph in-place, wherein the clustering comprises update steps, whereby at each update step, the sparse representation is modified in place to have nodes, clusters of original data points, and aggregated edges to preserve structure of the graph, the entity clustering engine comprising an interactive processor that receives feedback from an end user to the ETD data points comprising rules and constraints for merging, splitting and re-clustering the ETD data points and updates and revises clusters of ETD data points;
generate initial inclusion lists of those targeted entities that are determined to fulfill the audit request;
finalize the initial inclusion lists of targeted entities that fulfill the audit request to generate streamlined audit results; and
output the streamlined audit results to an end user to fulfill the input audit request.
20. A computer-implemented method of providing streamlined auditing of an audit request, comprising:
receiving an input audit request via the data interface; sourcing entity type data (ETD) from one or more databases;
mining ETD via an entity clustering engine to identify data point linkages amongst the ETD;
generating a graph of data using ETD data points and the mined linkages, where vertices of the graph comprise the ETD data points and N-dimensional edges of the graph comprise the mined data point linkages;
cluster the ETD data points using the mined data point linkages via a parallelized hardware component that updates a sparse representation of the ETD graph in-place, wherein the clustering comprises update steps, whereby at each update step, the sparse representation is modified in place to have nodes, clusters of original data points, and aggregated edges to preserve structure of the graph, the entity clustering engine comprising an interactive processor that receives feedback from an end user to the ETD data points comprising rules and constraints for merging, splitting and re-clustering the ETD data points and updates and revises clusters of ETD data points;
generating initial inclusion lists of those targeted entities that are determined to fulfill the audit request;
finalizing the initial inclusion lists of targeted entities that fulfill the audit request to generate streamlined audit results; and
outputting the streamlined audit results to an end user to fulfill the input audit request.