Computing system and methods for distributed generation and storage of complex relational data

• US 7,647,335 B1
• Filed: 08/30/2005
• Issued: 01/12/2010
• Est. Priority Date: 08/30/2005
• Status: Expired due to Fees

• Current Assignee: ATA SPA - Advanced Technology Assessment
• Original Assignee: ATA SPA - Advanced Technology Assessment
• Inventors: Colecchia, Federico
• Primary Examiner(s): NGUYEN, CAM LINH T
• Application Number:
  US11/216,243
• Time in Patent Office: 1,596 Days
• Field of Search: 707/1, 707/3, 707/10, 707/100, 707/101, 707/102, 707/103.X
• US Class Current: 1/1
• CPC Class Codes: G06F 16/26 Visual data mining; Browsing structured data
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

A system is disclosed for distributed generation and storage of relational data which can be represented as a network or sub-network. The system includes a plurality of distributed computing nodes (e.g., general purpose computers) configured with software for generating and storing relational data from input individual records. The system further includes a decentralized management subsystem for distribution of input records to the distributed computing nodes for processing (i.e., generation and storage of the relational data) and for providing access to the relational data to a user. The management subsystem includes a) a network management system process providing for dispatching input records to the nodes, b) an optimization process for optimizing the dispatch of input records to the nodes, c) an indexing process for indexing of records dispatched to the nodes and relational data generated at the nodes; and d) a query management process for responding to user queries and providing transparent access to the relational data.

Claims

1. A system for distributed generation, storage, and providing access to data structures corresponding to networks or sub-networks and representing said data structures in a graphical form via an algorithm that supports generation, manipulation and visualization of said graphical form, comprising:

a plurality of distributed computing nodes configured with software for generating said data structures corresponding to networks or sub-networks from input individual records, the data structures capable of being represented in a graphical form via an algorithm that supports generation, manipulation and visualization of said graphs; and

a decentralized software and hardware management subsystem for distribution of input records to the distributed computing nodes and for providing access to said data structures corresponding to networks or sub-networks to a user, comprising;

a) a network management system process coordinating dispatching input records or information associated therewith to the distributed computing nodes,

b) an optimization process for optimizing the dispatch of records or information associated therewith to the nodes,

c) an indexing process for indexing of records dispatched to the nodes or otherwise available to the nodes, and said data structures corresponding to networks or sub-networks generated at the nodes, processing local indexing data into overall indexes, and providing overall indexing data to the optimization process or other processes in the management subsystem as needed; and

d) a query management process for responding to queries of a user and providing transparent access to said data structures corresponding to networks or sub-networks for graphical display of said data structures to said user.

2. The system of claim 1, wherein the network management system process, optimization process, indexing process and query management process are implemented on one or more separate computing platforms, and wherein the decentralized distributed software and hardware system includes a mechanism for dynamically transferring the network management system process, optimization process, indexing process and query management process from one computing platform to another computing platform.

3. The system of claim 1, wherein the one or more computing platforms comprise distributed computing nodes in the plurality of distributed computing nodes.

4. The system of claim 1, wherein the input records comprise publications or portions thereof.

5. The system of claim 4, wherein the sub-network or network comprises a network selected from the group of networks consisting of a) an undirected network, and b) a directed network.

6. The system of claim 1, wherein the input records include relational data.

7. The system of claim 1, wherein the records comprise relational data and non-relational data, and wherein the network management process includes instructions for directing one of the distributed computing nodes to forward relational data associated with a record to another of the distributed computing nodes for processing.

8. The system of claim 1, wherein the query management process further supports the use of a visualization tool for generating a graphical visualization of the relational data requested by a user.

9. The system of claim 1, wherein the distributed computing nodes generate data structures corresponding to sub-networks, and wherein the query management process merges sub-networks from two or more distributed computing nodes in response to a user query.

10. The system of claim 1, wherein each of the distributed computing nodes includes an indexing process and wherein the output of the indexing process is provided to a computing platform executing the indexing process in the management subsystem.

11. The system of claim 1, wherein the management subsystem further comprises a mechanism for dynamically adding or removing a node to or from the system.

12. A method for distributed generation and storage of data structures corresponding to networks or sub-networks and representing said data structures in a graphical form via an algorithm that supports generation, manipulation and visualization of said graphical form, comprising the steps of:

receiving a collection of individual input records;

distributing the records to a plurality of individual, distributed computing nodes;

at the individual computing nodes, generating said data structures corresponding to networks or sub-networks and storing the data structures corresponding to networks or sub-networks as relational data locally, and

creating an index of the relational data stored at individual computing nodes to facilitate access to the relational data in response to a query of a user, the data structures capable of being represented in a graphical form via an algorithm that supports generation, manipulation and visualization of said graphical form,

wherein the relational data corresponding to said data structures for a set of multiple individual input records is generated and stored across a plurality of individual computing nodes; and

in response to said query providing said data structures to the user for graphical display of the data structures to the user.

13. The method of claim 12, wherein the method further comprises the steps of:

a) at the individual computing nodes, indexing local record data;

b) providing local record data indexed at the individual computing nodes to a computing platform executing an indexing process;

c) validating the indexing of the relational data generated at the individual computing nodes at the computing platform executing the indexing process; and

d) using the index of the relational data to select relational data to respond to the query from the user.

14. The method of claim 13, wherein the response to the user'"'"'s query is generated by a query management process accessing indexing data provided by the computing platform executing the indexing process.

15. The method of claim 12, wherein the step of distributing the records to a plurality of individual computing nodes is optimized according to at least one of (a) local computing resources available at the individual computing nodes, (b) pre-existing availability of the records or portions thereof at the individual computing nodes or storage devices accessible thereto, and (c) customizable optimization criteria.

16. The method of claim 15, wherein the customizable optimization criteria include at least one of (i) neighborhood coverage comprising information as to adjacent vertices in a network representation of the relational data and (ii) completeness, comprising relevant information for a given link in the network representation being stored at the same computing node.

17. The method of claim 15, wherein the step of distributing the records to a plurality of individual computing nodes is performed by a network management process executing in a first computing platform, the network management process coordinating distribution of the records to the plurality of individual computing nodes from optimization input from a records distribution optimization process.

18. The method of claim 17, wherein the network management process and the records distribution optimization process are executed on distinct computing platforms.

19. The method of claim 12, wherein the method further comprises the step of transferring at least one of the following processes from one computing platform to another computing platform:

a) a network management system process for coordinating dispatching records to the distributed computing nodes for processing,

b) an optimization process for optimizing the dispatch of records to the nodes for processing,

c) an indexing process of indexing of records dispatched to the nodes and relational data stored at the nodes; and

d) a query management process for responding to user queries of the distributed computing system and providing transparent access to the relational data.

20. The method of claim 12, wherein, in response to a user query, relational data from two or more individual computing nodes is merged into a merged representation of the relational data.

21. The method of claim 12, wherein the input records comprise publications or portions thereof.