Method for monitoring a decision-making process when pursuing an objective in economic, technical, and organizational fields

• US 7,849,036 B2
• Filed: 09/26/2007
• Issued: 12/07/2010
• Est. Priority Date: 10/17/2000
• Status: Expired due to Fees

• Current Assignee: Martine Naillon
• Original Assignee: Martine Naillon
• Inventors: Naillon, Martine
• Primary Examiner(s): Sparks; Donald
• Assistant Examiner(s): FERNANDEZ RIVAS, OMAR F
• Application Number:
  US11/861,563
• Publication Number:
  US 20080215521A1
• Time in Patent Office: 1,168 Days
• Field of Search: 706/10, 706/45, 706/62, 709/202, 714/25, 714/26, 714/37, 714/100
• US Class Current: 706/45
• CPC Class Codes: G06N 5/043 Distributed expert systems; Blackboards

Abstract

A method for controlling a decisional process when pursuing an aim in a specific application domain, such as economic, technical, and organizational fields. In the process, a system of agents is created, including creator cognitive agents of portable knowledge objects and constructor agents of knowledge objects by adding supplementary knowledge elements to the knowledge objects created. The system launches the decisional process by tracking the observation of an abnormality in the actual data of the application domain by the creation of an initial knowledge object representative of an alarm index and causes several decisional paths to be established linked to this object, under different points of view of different agents, each in its domain, and the decision is made following merging of decisional paths.

Claims

1. A method for making a decision when pursuing an overall aim, after discovery of an abnormality in an economic, technical, or organizational field, to counteract the abnormality, the method comprising:

associating with a collective knowledge manager a plurality of individual knowledge manager machines, each knowledge manager machine concerning a different point of view determined by the overall aim pursued;

establishing a structure of a first knowledge manager machine directly related to the field in which the abnormality has been detected and constituting a first decisional path, the structure of the first knowledge manager machine having a pyramidal shape, with a user manager at an apex, by associating user manager data processing agents located in a plurality of levels of the pyramidal shape, with,

in the lowermost of the levels, portable knowledge object creating agents watching actual data in a specific field and detecting abnormalities in the actual data of the first knowledge manager machine, each abnormality detected in the actual data that is being watched, causing in response, creation of a portable knowledge object as an index object, and

in the levels above the lowermost level including the knowledge object creating agents, knowledge object constructing agents adding additional knowledge elements to the portable knowledge objects created by the knowledge object creating agents;

transmitting the portable knowledge objects from the knowledge object constructing agents to the user manager and transmitting the portable knowledge object from the user manager to the collective knowledge manager;

selecting, in response to a request from the collective knowledge manager for further study of the portable knowledge objects constructed in the first knowledge manager machine, at least a second knowledge manger machine from the plurality of individual knowledge manager machines;

establishing a structure of the second knowledge manager machine, having a pyramidal shape, similar to the structure of the first individual knowledge manager machine, and constituting a second decisional path with a user manager at an apex, with knowledge object creating agents in a lowermost level of a plurality of levels in the pyramidal shape, and knowledge object constructing agents in levels above the knowledge object creating agents, to achieve the overall aim and results produced by the first individual knowledge manager machine; and

providing, in the second knowledge manager machine, the knowledge object creating agents watching actual data and detecting, events related to the knowledge objects created in the first individual knowledge manager machine and, upon detection of an event, creating a knowledge object which is transmitted to the user manager of the second individual knowledge manager machine and to the collective knowledge manager and constructed by addition of additional object elements in the knowledge object constructing agents of the second individual knowledge manager machine to enable the collective knowledge manager to merge results of the first and second decisional paths of the first and second individual knowledge manager machines, to enable a decision for counteracting the abnormality detected.

2. The method as claimed in claim 1, including

data processing agents organized into loops, with a data processing agent in one level of a pyramid of a master agent, and

at least one child agent in a level below the master agent so that a strategy followed by the master agent is an aim of the child agent, wherein the child agent is the master agent of at least one child agent in a level directly below the master agent.

3. A method for making a decision when pursuing an overall aim, after discovery of an abnormality in an economic, technical, or organizational field, to counteract the abnormality, the method comprising:

associating with a collective knowledge manager a plurality of individual knowledge manager machines, each knowledge manager machine concerning a different point of view determined by the overall aim pursued;

establishing a structure of a first knowledge manager machine directly related to the field and constituting a first decisional path, the structure of the first knowledge manager machine having a pyramidal shape, with a user manager at an apex, by associating user manager data processing agents located in a plurality of levels of the pyramidal shape, with,

in the lowermost of the levels, portable knowledge object creating agents watching actual data in a specific field and detecting abnormalities in the actual data, each abnormality causing creation of a portable knowledge object as an index object, and

in the levels above the lowermost level including the knowledge object creating agents, knowledge object constructing agents adding additional knowledge elements to the portable knowledge objects created by the knowledge object creating agents;

transmitting the portable knowledge objects from the knowledge object constructing agents to the user manager and transmitting the portable knowledge object from the user manager to the collective knowledge manager;

establishing a structure of a second of the individual knowledge manager machines, having a pyramidal shape, similar to the structure of the first individual knowledge manager machine, and constituting a second decisional path with a user manager at an apex, with knowledge object creating agents in a lower most level of a plurality of levels in the pyramidal shape, and knowledge object constructing agents in levels above the knowledge object creating agents, to achieve the overall aim and results produced by the first individual knowledge manager machine;

providing, in the second individual knowledge manager machine, the knowledge object creating agents watching actual data and detecting, events related to the knowledge objects created in the first individual knowledge manager machine and, upon detection of an event, creating a knowledge object which is transmitted to the user manager of the second individual knowledge manager machine and to the collective knowledge manager and constructed by addition of additional object elements in the knowledge object constructing agents of the second individual knowledge manager machine to enable the collective knowledge manager to merge results of the first and second decisional paths of the first and second individual knowledge manager machines, to enable a decision for counteracting the abnormality detected;

organizing data processing agents into loops, with a data processing agent in one level of a pyramid of a master agent;

locating at least one child agent in a level below the master agent so that a strategy followed by the master agent is an aim of the child agent, wherein the child agent is the master agent of at least one child agent in a level directly below the master agent; and

establishing a loop of the pyramid with a structuring agent comprising a service structuring agent and a data structuring agent, wherein the master agent of the loop obtains, from the service structuring agent, services considered in the service structuring agent to be necessary to accomplish the overall aim, and the strategy of the loop master agent, to create, for each service, a child agent, and to obtain from the data structuring agent, the data necessary for each child agent to implement an aim of the child agent.

4. The method as claimed in claim 3, wherein each knowledge object constructing agent is provided with aim and strategy rules.

5. The method as claimed in claim 4, wherein a knowledge object creating agent comprises aim and plan rules enabling the knowledge object creating agent to observe data concerning the field and to capture data representing an abnormality in the data concerning the field.

6. The method as claimed in claim 5, including a knowledge object created and provided with additional knowledge object elements, comprising a predetermined number of elements, each element concerning a knowledge aspect, indicating nature, context[H] and importance of the knowledge object.

7. The method as claimed in claim 6, wherein the user manager at the apex of the individual knowledge manager machines receives the knowledge object created, and including additional knowledge object elements to structure the knowledge objects and to transmit to the collective knowledge manager at least the knowledge object which has been created, as an index object, in the first individual knowledge manager machine of the field.

8. The method as claimed in claim 7, wherein the collective knowledge manager updates the knowledge object received by merging knowledge object versions of the knowledge object, received from each individual knowledge manager machine, to present to an operator the knowledge object in a newly structured configuration.

9. The method as claimed in claim 8, including producing a dialogue between the operator and the collective knowledge manager according to pre-established rules governing operation of the collective knowledge manager.

10. The method as claimed in claim 9, including displaying, by the collective knowledge manager, on an interface screen, configuration of the structured knowledge object.

11. The method as claimed in claim 9, including gathering, by the operator, more information on aspects of the knowledge object through the dialogue between the operator and the collective knowledge manager, according to predetermined rules, to implement a supplementary decision stage in which the collective knowledge manager sends a request to the individual knowledge manager machine that is most involved, thereby giving the information desired by the operator.

12. The method as claimed in claim 11, including, executing by the individual knowledge manager machine, a mission in response to a request by cooperating with at least one other individual knowledge manager machine according to predetermined rules.

13. The method as claimed in claim 12, wherein the individual knowledge manager machine which has received the request from the collective knowledge manager in the individual manager machine makes an investigation within the individual knowledge manager machine of the individual knowledge manager machine, in response to the request, in accordance with specific aim and strategy rules.

14. The method as claimed in claim 13, comprising a further supplementary decision stage, including a technical analysis of the knowledge object, completed by information items resulting from foregoing decision stages, enabling a decision to be made by the individual knowledge manager machine in accordance with specific aim and strategy rules.

15. The method as claimed in claim 8, wherein information items given by a knowledge object are represented in a structured knowledge tree, organized in knowledge classes, each knowledge object being detected as an index corresponding to a structured knowledge tree.