Back to results
US 7,480,712 B2Granted
Computer automated group detection
Assignee21st Century Technologies, Inc.
InventorMelanie Tina Moy
Life of this patent
Estimating the term…
Abstract
A method, system and computer program product implementing a computational algorithm that enables analysts to discover/find groups within a database of seemingly random electronic data in a relatively short period of time. A computer system is provided that includes a processor and program code executing on the processor for completing the following functions: (a) receiving a plurality of random electronic data corresponding to one or more activity within a monitored system; (b) discovering a group of associated data points from within the plurality of electronic data.
Claims31
In a data processing system, a method comprising:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationships between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- determining which identified groups among the one or more groups have overlaps greater than a pre-established maximum overlap;
- merging groups having overlaps greater than the pre-established maximum overlap to generated merged groups;
- and
- outputting a result set of groups including the merged groups and other potential groups for evaluation as groups of interest.
3 more elements
In a data processing system, a method comprising:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationships between the one or more entities;
- and
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria;
- generating a “seed” node list utilizing a pre-defined constraint parameter, wherein an eligible seed node can be pre-defined by an analyst as a node which has certain properties including, but not limited to, minimum degree (number of neighbors), minimum centrality, etc.;
- and
- linking neighbors that are “well-connected” to the seed node and neighbors that are linked to “many” other neighbors in the same set as an initial group;
- wherein a well-connected neighbor is pre-defined to be a neighbor node which shares at least X % as many edges with the seed node as the most well-connected neighbor, wherein X % is a selected one of a first default parameter or a first analyst-defined parameter, respectively selected based on execution of a first group detection algorithm or a second group detection algorithm;
- wherein the many other neighbors indicates that a node may only belong to the initial group if the node is connected to at least Y % of the seed node's well-connected neighbors, wherein Y % is a selected one of a second default parameter or a second analyst-defined parameter, respectively selected based on execution of a first group detection algorithm or a second group detection algorithm.
8 more elements
In a data processing system, a method comprising:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationships between the one or more entities:
- and
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria:
- analyzing relationship patterns among the individuals to determine when a group exists;
- and
- dynamically determining aliases for members within a potential groups of interest, wherein when an exception to a group of interest is identified such that every member is connected to every other member of the group with one exception, the exception is marked as one that is indicative of an alias, whereby the two members within the group of interest that are not connected are identified as potentially the same member having an alias identity within the group.
6 more elements
In a data processing system, a method comprising:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationships between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- identifying individual nodes that connect two groups together;
- tagging the individual node as a potential courier between the two groups;
- performing intra-spective analysis of the groups and nodes, including one or more of the following:
- (a) identifying and observing collaborations;
- (b) detecting and facilitating ad-hoc teams;
- and (c) identifying and integrating potentially isolated schools of thought;
- and
- performing extra-spective analysis of the groups and nodes, including one or more of the following:
- (a) identifying and tracking overt groups;
- (b) detecting and identifying covert groups;
- and (c) seeding further SNA (social network analysis) functions to categorize a group's activities.
10 more elements
In a data processing system, a method comprising:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationships between the one or more entities:
- and
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria;
- providing a focused set of entities for evaluating shared traits of an identified group of interest, including physical location, activities, and interests provided as attributes linked to respective nodes;
- and
- analyzing changes in a group structure over time and outputting said changes for evaluation.
6 more elements
A computer program product comprising:
- a computer readable storage medium;
- and
- program code on the computer readable storage medium that when executed by a processing device implements the following functions:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- determining which identified groups among the one or more groups have overlaps greater than a pre-established maximum overlap;
- merging groups having overlaps greater than the pre-established maximum overlap to generated merged groups;
- and
- outputting a result set of groups including the merged groups and other potential groups for evaluation as groups of interest.
6 more elements
A computer program product comprising:
- a computer readable storage medium;
- and
- program code on the computer readable medium that when executed by a processing device implements the following functions:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining communication relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria;
- generating a “seed” node list utilizing a pre-defined constraint parameter, wherein an eligible seed node can be pre-defined by an analyst as a node which has certain properties including, but not limited to, minimum degree (number of neighbors), minimum centrality, etc.;
- and
- linking neighbors that are “well-connected” to the seed node and neighbors that are linked to “many” other neighbors in the same set as an initial group;
- wherein a well-connected neighbor is pre-defined to be a neighbor node which shares at least X % as many edges with the seed node as the most well-connected neighbor, wherein X % is a selected one of a first default parameter or a first analyst-defined parameter on execution of a first group detection algorithm;
- wherein the many other neighbors indicates that a node may only belong to the initial group if the node is connected to at least Y % of the seed node's well-connected neighbors, wherein Y % is a selected one of a second default parameter or a second analyst-defined parameter on execution of a first group detection algorithm.
10 more elements
A computer program product comprising:
- a computer readable storage medium;
- and
- program code on the computer readable medium that when executed by a processing device implements the following functions:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining communication relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria;
- analyzing relationship patterns among the individuals to determine when a group exists;
- and
- dynamically determining aliases for members within a potential groups of interest, where in when an exception to a group of interest is identified such that every member is connected to every other member of the group with one exception, the exception is marked as one that is indicative of an alias, whereby the two members within the group of interest that are not connected are identified as potentially the same member having an alias identity within the group.
8 more elements
A computer program product comprising:
- a computer readable storage medium;
- and
- program code on the computer readable storage medium that when executed by a processing device implements the following functions:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- identifying individual nodes that connect two groups together;
- tagging the individual node as a potential courier between the two groups;
- performing intra-spective analysis of the groups and nodes, including one or more of the following:
- (a) identifying and observing collaborations;
- (b) detecting and facilitating ad-hoc teams;
- and (c) identifying and integrating potentially isolated schools of thought;
- and
- performing extra-spective analysis of the groups and nodes, including one or more of the following:
- (a) identifying and tracking overt groups;
- (b) detecting and identifying covert groups;
- and (c) seeding further SNA (social network analysis) functions to categorize a group's activities.
13 more elements
A computer program product comprising:
- a computer readable storage medium;
- and
- program code on the computer readable medium that when executed by a processing device implements the following functions:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- receiving the data set of entities within a network as an input graph of nodes representing each entity interconnected by edges defining communication relationships between two or more entities, said relationships including interactions, ownership, and trust;
- wherein each node and connected edges are provided attributes, which store the details of the node and the connected edges, such as an entity's identification (ID) and a time of occurrence for a specific interaction represented by one of the connected edges;
- performing said analyzing and identifying within the input graph, such that each of said one or more groups include a minimum number of nodes connected by edges that meet specific grouping criteria;
- providing a focused set of entities for evaluating shared traits of an identified group of interest, including physical location, activities, and interests provided as attributes linked to respective nodes;
- and
- analyzing changes in a group structure over time and outputting said changes for evaluation.
8 more elements
A data processing device comprising:
- a processor;
- a group detection utility that when executed on the processor provides the functions of:
- systematically analyzing a data set of entities within a network having one or more interconnections amongst the entities that indicates relationship patterns between the one or more entities;
- dynamically identifying one or more groups from within the data set that satisfy pre-established grouping criteria, which criteria collectively indicates specific relationships between the entities within a specific group relative to other entities within the specific group;
- determining which identified groups among the one or more groups have overlaps greater than a pre-established maximum overlap;
- merging groups having overlaps greater than the pre-established maximum overlap to generated merged groups;
- and
- outputting a result set of groups including the merged groups and other potential groups for evaluation as groups of interest.
5 more elements
Dependencies are inferred from the claim text.