Measuring node proximity on graphs with side information

• US 8,478,785 B2
• Filed: 12/15/2009
• Issued: 07/02/2013
• Est. Priority Date: 12/15/2009
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Jamjoom, Hani T., Qu, Huiming, Tong, Hanghang
• Primary Examiner(s): To, Baoquoc N
• Application Number:
  US12/638,531
• Publication Number:
  US 20110145262A1
• Time in Patent Office: 1,295 Days
• Field of Search: 707/790, 707/798
• US Class Current: 707/790
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence

Abstract

In a computerized data mining context, user input relating to positive and negative information is incorporated into node proximity measurements on a weighted, directed graph. Starting from a source node, links are added to nodes for which positive feedback is received. Where negative information is received, a sink node is substituted for nodes receiving negative information. Nodes neighboring that sink node have links added to the sink. These changes yield an altered graph. Afterwards, proximity information is determined from the altered graph.

Claims

1. A computer method comprising performing operations in at least one data processing device, the operations comprising:

embodying on at least one machine readable medium a representation of at least one graph representation of data, the representation comprising respective pluralities of nodes, links, and link weights;

receiving user input denoting positive and/or negative feedback with respect to at least one node in the graph;

altering at least one link and/or link weight in the embodiment of the graph, responsive to the feedback, in order to yield an altered graph; and

presenting a machine readable embodiment of a proximity value between a source and target node responsive to the altered graph.

2. The method of claim 1, wherein

the feedback is negative with respect to at least one node y;

and

the altering comprises

adding a sink node into the graph; and

for each negative node y;

finding neighbors of y;

adding a link from node y to the sink; and

adding a respective link from each neighboring node of node y to the sink.

3. The method of claim 1, wherein

the feedback is positive with respect to at least one node x;

and

altering comprises adding a link from the source node to each positive node x.

4. The method of claim 1, wherein presenting a proximity value comprises performing a random walk with restart.

5. The method of claim 1, wherein the operations further comprise presenting the proximity value as a ranking of content to a user.

6. The method of claim 1, wherein the proximity value comprises a representation of a relationship between content.

7. The method of claim 1, wherein the graph is a directed graph and the links have direction.

9. The system of claim 1, wherein

the feedback is negative with respect to at least one node y;

and

the altering comprises

adding a sink node into the graph; and

for each negative node y;

finding neighbors of y;

adding a link from node y to the sink; and

adding a respective link from each neighboring node of node y to the sink.

8. A system comprising:

at least one data processing device;

at least one network and/or user interface device for communicating with the data processing device;

at least one medium for embodying at least machine executable code and data in machine readable form;

the code comprising instructions for causing the data processing device to perform operations on the data, the operations comprising

embodying on at least one machine readable medium a representation of at least one graph representation of data, the representation comprising respective pluralities of nodes, links, and link weights;

receiving user input denoting positive and/or negative feedback with respect to at least one node in the graph;

altering at least one link and/or link weight in the embodiment of the graph, responsive to the feedback, in order to yield an altered graph; and

presenting a machine readable embodiment of a proximity value between a source and target node responsive to the altered graph.

10. The system of claim 8, wherein

the feedback is positive with respect to at least one node x;

and

altering comprises adding a link from the source node to each positive node x.

11. The system of claim 8, wherein presenting a proximity value comprises performing a random walk with restart.

12. The system of claim 8, wherein the operations further comprise presenting the proximity value as a ranking of content to a user.

13. The system of claim 8, wherein the proximity value comprises a representation of a relationship between content.

14. A computer program product for performing operations, the computer program product comprising a storage medium readable by a processing circuit and storing instructions to be run by the processing circuit for performing a method comprising:

embodying on at least one machine readable medium a representation of at least one graph representation of data, the representation comprising respective pluralities of nodes, links, and link weights;

receiving user input denoting positive and/or negative feedback with respect to at least one node in the graph;

altering at least one link and/or link weight in the embodiment of the graph, responsive to the feedback, in order to yield an altered graph; and

presenting a machine readable embodiment of a proximity value between a source and target node responsive to the altered graph.

15. The program product of claim 14, wherein

the feedback is negative with respect to at least one node y;

and

the altering comprises

adding a sink node into the graph; and

for each negative node y;

finding neighbors of y;

adding a link from node y to the sink; and

adding a respective link from each neighboring node of node y to the sink.

16. The program product of claim 14, wherein

the feedback is positive with respect to at least one node x;

and

altering comprises adding a link from the source node to each positive node x.

17. The program product of claim 14, wherein presenting a proximity value comprises performing a random walk with restart.

18. The program product of claim 14, wherein the operations further comprise presenting the proximity value as a ranking of content to a user.

19. The program product of claim 14, wherein the proximity value comprises a representation of a relationship between content.