System and method for networked decision making support

• US 8,433,711 B2
• Filed: 09/09/2005
• Issued: 04/30/2013
• Est. Priority Date: 09/09/2005
• Status: Active Grant

• Current Assignee: Rateze Remote Mgmt. L.L.C. (Intellectual Ventures LLC)
• Original Assignee: Kang Jo Mgmt LLC (Intellectual Ventures LLC)
• Inventors: Odom, Paul S., Way, Robert D.
• Primary Examiner(s): Perveen, Rehana
• Assistant Examiner(s): Wong, Huen
• Application Number:
  US11/222,046
• Publication Number:
  US 20070061128A1
• Time in Patent Office: 2,790 Days
• Field of Search: None
• US Class Current: 707/748
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems

Abstract

The present invention provides systems and methods for decision making support. The decision making support can be used, for example, for help desk troubleshooting, medical diagnosis, financial planning, and expert device. The decision support system of the invention includes a processor and a memory having a decision support program module, a phase dictionary, one or more solution nodes and optionally a topic identification information module. The decision support module is configured to enable the processor to receive a problem constraint and one or more user-specified phrases, which are then used to search nodes in a solution domain to identify a set of nodes potentially related to the user'"'"'s problem or question. The set of nodes potentially related to the user'"'"'s problem or question are ranked and displayed. Methods to determine potential solution nodes and to update solution nodes within the decision support system are provided.

Claims

1. A method for identifying a set of related diagnoses and questions related to a problem in a solution domain having a plurality of nodes, comprising:

receiving a constraint at a computing device;

searching, by the computing device, the plurality of nodes to generate a first set of related nodes, wherein each node in the first set of related nodes includes;

at least a portion of the constraint;

a phrase comprising a diagnosis, a problem statement, or a question;

a response to the phrase;

a type indicator indicating whether the phrase is a diagnosis, problem statement, or a question; and

a success counter configured to store a running count of the number of times the respective node has been chosen as a solution to the problem;

generating a first phrase list at the computing device, wherein the first phrase list includes one or more of the phrases contained within at least some of the first set of related nodes, each phrase in the first phrase list being selectable;

receiving, at the computing device, a selection of one or more phrases from the first phrase list; and

searching, by the computing device, the plurality of nodes to generate a second set of related nodes, wherein each node in the second set of related nodes includes any of the one or more phrases selected, and wherein the success counter is used to weight a ranking of the respective node in the second set of related nodes.

2. The method of claim 1, wherein the first phrase list comprises a first phrase to include a question relating to the constraint, a second phrase to include a diagnosis relating to the constraint, and a third phrase to include a problem statement relating to the constraint.

3. The method of claim 1, wherein each node includes a plurality of responses to information presented in the node.

4. The method of claim 3, wherein the one or more phrases include at least one of the plurality of responses in a node.

5. The method of claim 1, wherein each node further comprises:

a short description;

an author identification;

a title; and

one or more topics that relate to the node.

6. The method of claim 5, wherein generating a first phrase list comprises:

retrieving the one or more phrases from a node in the first set of related nodes to generate a first set of phrases;

removing redundant phrases from the first set of phrases to generate a second set of phrases; and

ordering the second set of phrases.

7. The method of claim 6, wherein ordering the second set of phrases comprises ordering the phrases alphabetically.

8. The method of claim 6, further comprising separating the second set of phrases into one or more categories.

9. A method for identifying a set of diagnoses and questions related to a problem in a solution domain having a plurality of nodes, comprising:

receiving, at a computing device, a constraint comprising one or more phrases;

searching, by the computing device, the plurality of nodes to generate a first set of related nodes, wherein the search selects nodes that include the constraint;

generating a phrase list at the computing device, wherein the phrase list includes one or more phrases contained within the first set of related nodes, each phrase being selectable;

receiving, at the computing device, a selection of one or more phrases from the first phrase list;

removing, by the computing device, nodes from the first set of related nodes that include an inverse of the phrases selected from the first phrase list to generate a second set of related nodes; and

ranking, by the computing device, the nodes in the second set of related nodes, wherein ranking the nodes in the second set of related nodes comprises;

determining, by the computing device, the number of phrases that match any of the one or more phrases;

determining, by the computing device, the number of phrases that do not match any of the one or more phrases;

multiplying, by the computing device, the number of matched phrases by a first weight to generate a matched value;

multiplying, by the computing device, the number of unmatched phrases by a second weight to generate an unmatched value;

subtracting, by the computing device, the unmatched value from the matched value to generate a weighted score for the node; and

incrementing, by the computing device, a success counter associated with a node when the node is identified as a proper diagnosis of the problem, wherein the success counter is used to further weight the ranking of the respective node in the second set of related nodes.

10. The method of claim 9, wherein ranking the nodes in the second set of related nodes further comprises ranking the nodes in the second set of related nodes in ascending or descending order of weighted scores.

11. The method of claim 9, further comprising:

adjusting the ranking of the nodes in the second set of related nodes using a success counter associated with each node in the second set of related nodes.

12. The method of claim 9, further comprising:

prior to receiving a constraint generating topics associated the plurality of nodes in the solution domain, wherein a topic is a word or word combination that occurs a statistically significant number of times in a node.

13. The method of claim 12, further comprising:

determining whether the constraint includes a word combination that is a topic.

14. The method of claim 13, wherein if the constraint includes a word combination that is a topic, searching the plurality of nodes to generate the first set of related nodes comprises selecting nodes that include the words in the word combination of the constraint contiguously.

15. The method of claim 13, wherein if the constraint includes a word combination that is a topic, searching the plurality of nodes to generate the first set of related nodes comprises selecting nodes that include the words in the word combination of the constraint separated by a specified number of words.

16. The method of claim 12, further comprising:

determining the frequency that each topic occurs within a node.

17. The method of claim 16, further comprising:

if the second set of nodes is empty, comparing each word and word combination in the constraint to the topics;

determining the frequency of occurrence for each word and word combination in the constraint;

modifying the constraint to remove the word or word combination having the lowest frequency of occurrence; and

searching the plurality of nodes to generate a third set of related nodes.

18. The method of claim 9, further comprising:

if the second set of related nodes is empty, investigating the problem off-line to identify a diagnosis for the problem;

entering a new node to the domain, wherein the new node includes the diagnosis identified for the problem; and

integrating the new node into the domain.

19. A computer-readable medium having instructions thereon that, if executed by a computing device, cause the computing device to perform operations comprising:

receiving a constraint;

searching the plurality of nodes to generate a first set of related nodes, wherein each node in the first set of related nodes includes;

at least a portion of the constraint;

a phrase comprising a diagnosis, a problem statement, or a question;

a response to the phrase;

a type indicator indicating whether the phrase is a diagnosis, problem statement, or a question; and

a success counter configured to store a running count of the number of times the respective node has been chosen as a solution to the problem;

generating a first phrase list, wherein the first phrase list includes one or more of the phrases contained within at least some of the first set of related nodes, each phrase in the first phrase list being selectable;

receiving a selection of one or more phrases from the first phrase list; and

searching the plurality of nodes to generate a second set of related nodes, wherein each node in the second set of related nodes includes any of the one or more phrases selected, and wherein the success counter is used to weight a ranking of the respective node in the second set of related nodes.

20. A computer-readable medium having instructions thereon that, if executed by a computing device, cause the computing device to perform operations comprising:

receiving a constraint comprising one or more phrases;

searching the plurality of nodes to generate a first set of related nodes, wherein the search selects nodes that include the constraint;

generating a phrase list, wherein the phrase list includes one or more phrases contained within the first set of related nodes, each phrase being selectable;

receiving a selection of one or more phrases from the first phrase list;

removing nodes from the first set of related nodes that include an inverse of the phrases selected from the first phrase list to generate a second set of related nodes; and

ranking the nodes in the second set of related nodes, wherein ranking the nodes in the second set of related nodes comprises;

determining the number of phrases that match any of the one or more phrases;

determining the number of phrases that do not match any of the one or more phrases;

multiplying the number of matched phrases by a first weight to generate a matched value;

multiplying the number of unmatched phrases by a second weight to generate an unmatched value;

subtracting the unmatched value from the matched value to generate a weighted score for the node; and

incrementing a success counter associated with a node when the node is identified as a proper diagnosis of the problem, wherein the success counter is used to further weight the ranking of the respective node in the second set of related nodes.