Discovering communities-of-practice

1. A computer-implemented method for discovering communities-of-practice from a communications network structure for a set of individual members, the method comprising:

selecting a current member of said, set;

deriving from the communications network data structure represented as a database including data for a given characteristic for each member of the set at least one communities-of-practice which identifies a group of members with a common interest or goal within said network structure among individual members coupled to said current member by deconstructing said network data structure associated with said current member based in part upon the given characteristic relational to said communications network data structure;

representing each of the at least one derived communities-of-practice in a respective data structure; and

storing each communities-of-practice data structure in a memory.

2. The computer-implemented method as set forth in claim 1 further comprising:

selecting a succession of members of said set,

repeating said deriving of at least one communities-of-practice data structure identifying a group of members with a common interest or goal for each of said members, and

deriving a plurality of communities-of-practice data structures for said communications network data structure.

3. The computer-implemented method as set forth in claim 2 further comprising:

repeating said deriving for a first set of arbitrarily selected members,

deriving a first plurality of communities-of-practice data structures based on said first set,

selecting at least one second set of arbitrarily selected members,

repeating said deriving a second plurality of communities-of-practice data structures based on said at least one second set,

based on the data stored in the communities-of-practice data structures from said deriving for each said set of members,

forming a communities-of-practice data structure for all said members.

4. The computer-implemented method as set forth in claim 1 wherein said communications network data structure is based on addressing data of communications by said set of individual members.

5. The computer-implemented method as set forth in claim 1 wherein said given characteristic is a betweenness centrality function.

6. A computer-implemented method for discovering communities-of-practice based upon a communications network construct data structure associated with a set of members, the method comprising:

selecting a first member represented in the construct data structure;

determining one or more communities-of-practice for association with the first member which identify other members having at least one common interest or goal with the first member by distinguishing and suppressing isolated inter-community communication linkages within the construct data structure;

representing each of the one or more identified communities-of-practice in a respective data structure; and

storing each communities-of-practice data structure in a memory.

7. The computer-implemented method as set forth in claim 6 wherein said selecting a first member of the construct is based on betweenness centrality determination.

8. The computer-implemented method as set forth in claim 7 wherein said betweenness centrality determination further comprises:

distinguishing inter-communities-of-practice linkages which link a predetermined number of members of the construct data structure;

defining a relatively high betweenness characteristic, from intra-community linkages between members, and

defining a relatively low betweenness characteristic.

9. The computer-implemented method as set forth in claim 6 said distinguishing and suppressing isolated inter-community communication linkages within the construct data structure further comprising:

repeatedly identifying inter-community linkages of a predetermined large betweenness characteristic, and

removing said biter-community linkages.

10. A computer-implemented method of doing business, the method comprising:

receiving pairwise communications addressing data;

deriving a hierarchical relationship of individuals based upon said addressing data;

representing said hierarchical relationship in a data structure stored in a memory;

deconstructing said hierarchical relationship as represented in the data structure by deriving communities-of-practice among said individuals based upon a given characteristic relational to said addressing data wherein a community-of-practice represents a group of members sharing a common interest or goal; and

representing each said derived communities-of-practice in a respective data structure;

storing each communities-of-practice data structure in the memory;

and traversing one of the communities-of-practice data structure associated with a common interest or goal in order to identify the group of members associated with that common interest or goal.

11. The computer-implemented method as set forth in claim 10 wherein said given characteristic relational to said addressing data is a betweenness centrality determination.

12. The computer-implemented method as set forth in claim 11 wherein an organization is represented by a plurality of N-nodes, where N is the sum total of all nodes, connected by edges, and a single edge connecting a leaf node has a betweenness of minus-one, −1, containing a shortest path from that leaf node all −1 other nodes, and a stopping criterion is set for component communities-of-practice of less than and equal to a minimum number of nodes forming a communities-of-practice is the highest betweenness of any edge in any component is equal to or less than N−1.

13. A computer-implemented system for discovering communities-of-practice, the system comprising:

means for forming a communications network data structure based upon addressing information for pairwise communications over a predetermined period of time;

means for selectively disassembling said communications network data structure based in part upon distinguishing and suppressing inter-community communications linkages as represented in the network data structure such that correspondents of said inter-community communications are assigned to different communities-of-practices;

means for representing each of the different communities-of-practice in a respective data structure; and

storing each communities-of-practice data structure in a memory.

14. The computer-implemented system as set forth in claim 13 , said means for selectively disassembling further comprising:

means for identifying inter-community communication linkage representations among members of said structure.

15. The computer-implemented system as set forth in claim 14 , said means for identifying inter-community communications linkage representations further comprises:

means for determining betweenness centrality of said communication linkage representations.

16. The computer-implemented system as set forth in claim 15 , said means for determining betweenness centrality of said communication linkage representations further comprises:

means for selecting each member associated with the communications network data structure sequentially,

means for finding shortest linkage paths as represented in the communications network data structure between a current selected member and all other members,

means for assigning a betweenness value to each of said shortest linkage paths,

means for tabulating a running total for each of said linkage paths, and

assigning a betweenness centrality value to each of said linkage paths equal to one-half the running total therefor.

17. The computer-implemented system as set faith in claim 16 , said means for selecting each member further comprising:

means for arbitrarily selecting members up to a predetermined number of members.

18. The computer-implemented system as set forth in claim 16 comprising:

means for selecting a current member of the communications network data structure;

means for forming a first subset of members of the communications network data structure including said current member and each member of the network having a threshold number of communication links with said current member;

means for finding a characteristic associating members within said first subset such that said first subset is divided into communities-of-practice based on said characteristic;

means for storing a respective data structure representing each communities-of-practice; and

means for iterating forming subsets and dividing said subsets into communities-of-practice.

19. The computer-implemented system as set forth in claim 18 further comprising:

means for aggregating result by said means for iterating and for providing a composite communities-of-practice structure therefrom.

20. A computer readable medium having encoded thereon instructions which when executed by a processor perform a method for discovering communities-of-practice from a communications network data structure, the method comprising:

selecting a first member represented in the communications network data structure;

determining one or more communities-of-practice structures for association with the first member which identity other members having at least one common interest or goal with the first member by distinguishing and suppressing isolated inter-community communication linkages within the communications network data structure;

representing each of the one or more identified communities-of-practice in a respective data structure; and

storing each communities-of-practice data structure in a memory.