Generating and evaluating expert networks

• US 8,990,058 B2
• Filed: 09/12/2012
• Issued: 03/24/2015
• Est. Priority Date: 09/11/2012
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Cheng, Winnie W., Miao, Gengxin, Moulic, James R., Tao, Shu
• Primary Examiner(s): Thangavelu, Kandasamy
• Application Number:
  US13/611,530
• Publication Number:
  US 20140074443A1
• Time in Patent Office: 923 Days
• Field of Search: 703/2, 703/6, 703/13, 706/12, 706/46, 706/52
• US Class Current: 703/6
• CPC Class Codes: H04L 41/0816 the condition being an adaptation, e.g. in response to network events
• CPC Class Codes: H04L 41/142 using statistical or mathematical methods
• CPC Class Codes: H04L 41/145 involving simulating, designing, planning or modelling of a network
• CPC Class Codes: H04L 41/22 comprising specially adapted graphical user interfaces [GUI]
• CPC Class Codes: Y04S 40/00 Systems for electrical power generation, transmission, distribution or end-user application management characterised by the use of communication or information technologies, or communication or information technology specific aspects supporting them

Abstract

A method comprises obtaining a model of at least one collaborative network, modeling at least one routing pattern for one or more nodes of a given collaborative network, simulating one or more virtual network topologies of the given collaborative network using the at least one model and the at least one routing pattern, and configuring a given one of the one or more virtual network topologies on the given collaborative network. The given collaborative network comprises a plurality of nodes and one or more edges between pairs of the plurality of nodes. At least one of the obtaining, modeling, simulating and configuring steps are performed by a processor device.

Claims

1. A method, comprising:

obtaining, using at least one processor, at least one model of at least one collaborative network;

modeling, using the at least one processor, at least one routing pattern for one or more nodes of a given collaborative network;

simulating, using the at least one processor, one or more virtual network topologies of the given collaborative network using the at least one model and the at least one routing pattern; and

configuring, using the at least one processor, a given one of the one or more virtual network topologies on the given collaborative network based on simulation results;

wherein the given collaborative network comprises a plurality of nodes and one or more edges between pairs of the plurality of nodes.

2. The method of claim 1, wherein the at least one model is a graphical model of the at least one human network.

3. The method of claim 1, wherein the step of obtaining at least one model comprises generating the at least one model based at least in part on historical data of at least one collaborative network.

4. The method of claim 1, wherein the step of obtaining at least one model comprises retrieving the at least one model from a database.

5. The method of claim 1, wherein the one or more virtual network topologies are simulated to optimize one or more network performance metrics of the given collaborative network.

6. The method of claim 5, wherein the one or more network performance metrics comprise at least one of:

an in-degree distribution of a number of incoming edges for each of the one or more nodes of the given collaborative network, an incoming edge representing a connection which is received at a node;

an out-degree distribution of a number of outgoing edges for each of the one or more nodes of the given collaborative network, an outgoing edge representing a connection which originates at a node; and

a clustering coefficient measuring a number of edges for a given node in the given collaborative network.

7. The method of claim 6, wherein the clustering coefficient is a function of at least one of:

similarity between expertise of nodes in the given network and interactions between nodes in the given network.

8. The method of claim 1, wherein the step of modeling at least one routing pattern further comprises identifying the number of edges required to route information between nodes in the given collaborative network until the information is received by a destination node.

9. The method of claim 8, wherein the information represents an unsolved problem and the destination node is a node capable of solving the unsolved problem.

10. The method of claim 8, wherein identifying the number of edges required to route information comprises:

reading historical routing data from a database; and

creating stochastic models to emulate routing behavior of nodes;

wherein the emulated routing behavior is based at least in part on actual route information observed from historical data.

11. The method of claim 1, wherein the given one of the one or more virtual network topologies is configured subject to one or more physical constraints of the given collaborative network.

12. The method of claim 11, wherein the one or more physical constraints comprises at least one of an organizational hierarchy of nodes in the given collaborative network and a geographic distribution of nodes in the given collaborative network.

13. The method of claim 1, wherein the one or more virtual network topologies are simulated subject to one or more physical constraints of the given collaborative network.

14. The method of claim 1, wherein the step of simulating one or more virtual network topologies further comprises:

evaluating a routing performance of the one or more virtual network topologies;

wherein the routing performance is based at least in part on the number of edges required to route information between nodes in the given collaborative network until the information is received by a destination node.

15. The method of claim 14, wherein the step of simulating one or more virtual network topologies further comprises identifying a recommended virtual network topology based at least in part on the evaluated routing performance of the one or more virtual network topologies.

16. The method of claim 1, wherein the step of simulating one or more virtual network topologies further comprises generating at least one performance report comprising quality metrics for at least one of the one or more virtual network topologies.

17. The method of claim 16, wherein the quality metrics comprise at least one of:

a number of edges required to route each of a plurality of tasks;

a number of nodes required to route each of a plurality of tasks;

an estimated time required to complete each of a plurality of tasks; and

an estimated number of unsolved tasks in a plurality of tasks.

18. The method of claim 1, wherein the given collaborative network is a support network, and each of the plurality of nodes represents a human expert user of the support network.

19. The method of claim 1, wherein a given one of the one or more virtual network topologies is simulated based at least in part on a power-law distribution of edges among a plurality of nodes in the given network topology.

20. The method of claim 1, wherein the step of simulating one or more virtual network topologies comprises:

randomly placing a plurality of nodes on a two-dimensional rectangular space;

adding periodic boundary conditions in the two-dimensional rectangular space;

determining an awareness coefficient and an exposure coefficient for each of the plurality of nodes;

creating an edge between a first node and a second node when the distance between first node and the second node is less than a product of the awareness coefficient of the first node and the exposure coefficient for the second node; and

swapping one or more edges between pairs of nodes which cross a boundary condition.

21. The method of claim 20, wherein the awareness coefficient is a function of a given node'"'"'s knowledge of one or more characteristics of one or more of the plurality of nodes and wherein the exposure coefficient is a function of the plurality of nodes'"'"' knowledge of one or more characteristics of the given node.