System and method for graph pattern analysis
In some example embodiments, a system and method are provided for graph pattern analysis. In example embodiments, pattern data of a primary network that includes data relating to relationships between entities are received. A reference matrix comprising a plurality of secondary network determined based on the pattern data of the primary network is generated. A graphical display of the primary network and the reference matrix are displayed on a user interface. A selection of a secondary network from the plurality of secondary networks of the reference matrix is received. The selected secondary network has similar matching characteristics with at least a portion of the primary network. In response to the selection, the primary network is classified as a classification type related to the selected secondary network.
1. A method, comprising:
performing, by a computer system, a training operation to train an artificial neural network (ANN) based on a historical dataset;
accessing, by the computer system, pattern data specifying a plurality of characteristics of a primary graph, wherein the primary graph corresponds to a primary network that includes a plurality of entities, and wherein the pattern data indicates:
a plurality of nodes corresponding to the plurality of entities in the primary network; and
a plurality of edges forming connections between the plurality of nodes, wherein the connections are indicative of a relationship between the plurality of entities;
generating, by the computer system, a plurality of secondary graphs that are based on the primary graph, wherein a particular secondary graph, of the plurality of secondary graphs, is associated with a particular classification type and includes a second plurality of edges connecting a second plurality of nodes, wherein, for the particular secondary graph, the generating includes:
parsing the pattern data to identify one or more of the plurality of characteristics of the primary graph; and
applying an input value based on the one or more characteristics to the ANN to generate the particular secondary graph;
sending, by the computer system, information indicative of the plurality of secondary graphs to a client device to graphically depict the plurality of secondary graphs via a graphical user interface on the client device;
receiving, by the computer system from the client device, a selection of the particular secondary graph via the graphical user interface, wherein the selection indicates a similarity between at least a portion of the primary graph and the particular secondary graph; and
based on the selection of the particular secondary graph, classifying, by the computer system, the primary graph as being associated with the particular classification type.
2. The method of claim 1 , further comprising:
storing, by the computer system, at least a portion of the pattern data corresponding to the primary graph for future use as a secondary graph having the particular classification type.
3. The method of claim 1 , wherein the parsing the pattern data includes using a retrieved parsing grammar to identify the one or more of the plurality of characteristics of the primary graph.
4. The method of claim 1 , wherein the primary network corresponds to a suspected fraud network such that a plurality of accounts of entities in the suspected fraud network are represented by the plurality of nodes in the primary graph.
5. The method of claim 4 , wherein one or more transactions between the one or more accounts of entities in the suspected fraud network are represented by the plurality of edges in the primary graph.
6. The method of claim 1 , wherein the primary network corresponds to a marketing network such that a plurality of entities in the marketing network are represented by the plurality of nodes in the primary graph.
7. The method of claim 1 , wherein the pattern data is specified using extensible markup language (XML) format.
8. A system, comprising:
at least one processor;
a non-transitory, computer-readable medium having instructions stored thereon that are executable by the at least one processor to cause the system to:
perform a training operation to train an artificial neural network (ANN) based on a historical dataset;
access pattern data specifying a plurality of characteristics of a primary graph, wherein the primary graph corresponds to a primary network that includes a plurality of entities, and wherein the pattern data indicates:
a plurality of nodes corresponding to the plurality of entities in the primary network; and
a plurality of edges forming connections between the plurality of nodes, wherein the connections are indicative of a relationship between the plurality of entities;
generate a plurality of secondary graphs that are based on the primary graph, wherein a particular secondary graph, of the plurality of secondary graphs, is associated with a particular classification type and includes a second plurality of edges connecting a second plurality of nodes, wherein, for the particular secondary graph, the generating includes:
parsing the pattern data to identify one or more of the plurality of characteristics of the primary graph; and
applying an input value based on the one or more characteristics to the ANN to generate the particular secondary graph;
send information indicative of the plurality of secondary graphs to a client device to graphically depict the plurality of secondary graphs via a graphical user interface on the client device;
receive, from the client device, a selection of the particular secondary graph via the graphical user interface, wherein the selection indicates a similarity between at least a portion of the primary graph and the particular secondary graph; and
based on the selection of the particular secondary graph, classify the primary graph as being associated with the particular classification type.
9. The system of claim 8 , wherein the instructions are further executable to cause the system to:
store at least a portion of the pattern data corresponding to the primary graph for future use as a secondary graph having the particular classification type.
10. The system of claim 8 , wherein parsing the pattern data includes using a retrieved parsing grammar to identify the one or more of the plurality of characteristics of the primary graph.
11. The system of claim 8 , wherein the pattern data is specified using comma-separated values (CSV) format.
12. The system of claim 8 , wherein the primary network corresponds to a suspected fraud network such that a plurality of accounts of entities in the suspected fraud network are represented by the plurality of nodes in the primary graph.
13. The system of claim 12 , wherein one or more transactions between the one or more accounts of entities in the suspected fraud network are represented by the plurality of edges in the primary graph.
14. The system of claim 8 , wherein the primary network corresponds to a marketing network such that a plurality of entities in the marketing network are represented by the plurality of nodes in the primary graph.
15. A non-transitory, machine-readable medium having instructions stored thereon that are executable by a machine to perform operations comprising:
performing a training operation to train an artificial neural network (ANN) based on a historical dataset;
accessing pattern data specifying a plurality of characteristics of a primary graph, wherein the primary graph corresponds to a primary network that includes a plurality of entities, and wherein the pattern data indicates:
a plurality of nodes corresponding to the plurality of entities in the primary network; and
a plurality of edges forming connections between the plurality of nodes, wherein the connections are indicative of a relationship between the plurality of entities;
generating a plurality of secondary graphs that are based on the primary graph, wherein a particular secondary graph, of the plurality of secondary graphs, is associated with a particular classification type and includes a second plurality of edges connecting a second plurality of nodes, wherein, for the particular secondary graph, the generating includes:
parsing the pattern data to identify one or more of the plurality of characteristics of the primary graph; and
applying an input value based on the one or more characteristics to the ANN to generate the particular secondary graph;
sending information indicative of the secondary graphs to a client device to graphically depict the plurality of secondary graphs via a graphical user interface on the client device;
receiving, from the client device, a selection of the particular secondary graph via the graphical user interface, wherein the selection indicates a similarity between at least a portion of the primary graph and the particular secondary graph; and
based on the selection of the particular secondary graph, classifying the primary graph as being associated with the particular classification type.
16. The non-transitory, machine-readable medium of claim 15 , wherein the operations further comprise storing at least a portion of the pattern data corresponding to the primary graph for future use as a secondary graph having the particular classification type.
17. The non-transitory, machine-readable medium of claim 15 , wherein parsing the pattern data includes using a retrieved parsing grammar to identify the one or more of the plurality of characteristics of the primary graph.
18. The non-transitory, machine-readable medium of claim 15 , wherein the pattern data includes a Joint Photographic Experts Group (JPEG) file depicting the primary network.
19. The non-transitory, machine-readable medium of claim 15 , wherein the primary network corresponds to a suspected fraud network such that a plurality of accounts of entities in the suspected fraud network are represented by the plurality of nodes in the primary graph.
20. The non-transitory, machine-readable medium of claim 15 , wherein the information indicative of the secondary graphs is specified in an XML file.