Application integration system and method using intelligent agents for integrating information access over extended networks

• US 20050027495A1
• Filed: 05/24/2004
• Published: 02/03/2005
• Est. Priority Date: 10/03/2000
• Status: Active Grant

• Current Assignee: Celcorp Incorporated
• Original Assignee: Celcorp Incorporated
• Inventors: Matichuk, Bruce
• Granted Patent: US 7,454,399 B2
• Time in Patent Office: Days
• US Class Current: 703/2
• CPC Class Codes: G06F 9/45512 Command shells

Abstract

Models that encapsulate information on how remote applications are controlled and data is accessed are created to be used by a runtime agent that intelligently reasons out goal-oriented plans and execute tasks to solve problem statements generated by a user interface or other software agents. The models may be used to assemble workflows using a reasoning process. In particular, the models of the present invention include a state that includes a planning state and a navigational value. In embodiments of the invention, the models may include intelligent agents, a loop state, submodel invocation that allows the models to invoke agents over a network, and the ability to integrate the three application layers in a computer.

Claims

1. A computer system navigation planner, said computer system comprising:

a processor;

a computer memory coupled to said processor;

one or more computer application models stored in said computer memory, each of said models comprising a state comprising an action and one or more Boolean expressions; and

a navigation planner stored in said computer memory;

wherein when said navigation planner receives a problem statement comprising an initial condition and an end condition, each of said conditions comprising one or more Boolean expressions, said navigation planner accessing said computer application models to create a workflow of solving said problem statement, said workflow comprising an ordered combination of said models such that said ordered combination of models contain Boolean expressions that combine to satisfy said initial condition Boolean expression and end condition Boolean expression.

2. The computer system navigation planner of claim 1 wherein said state defines a loop that represents a series of repeated steps according a Boolean expression associated with said loop state.

3. The computer system navigation planner of claim 1 wherein said state defines a pattern-type used in screen navigation wherein a start of an associated action is determined at run time by evaluating an on-screen pattern.

4. The computer system navigation planner of claim 1 wherein said state represents plural services which are loosely connected.

5. The computer system navigation planner of claim 4 wherein said state a first action associated with said state invokes a second action that shares a knowledge base, such that knowledge accessed by said second action is not accessed by said first action.

6. A computer implemented navigation planning method comprising:

a receiving a problem statement comprising an initial condition and an end condition, each of said conditions comprising one or more Boolean expressions;

defining one or more computer application models stored in computer memory, each of said models comprising a state comprising an action and one or more Boolean expressions; and

creating a workflow for solving said problem statement, said workflow comprising an ordered combination of said models such that said ordered combination of models contain Boolean expressions that combine to satisfy said initial condition Boolean expression and said end condition Boolean expression.

7. The planning method of claim 6 wherein said state defines a loop that represents a series of repeated steps according a Boolean expression associated with said loop state.

8. The planning method of claim 6 wherein said state defines a pattern-type used in screen navigation wherein a start of an associated action is determined at run time by evaluating an on-screen pattern.

9. The planning method of claim 6 wherein said state represents plural actions which are loosely connected.

10. The planning method of claim 9 wherein said state a first action associated with said state invokes a second action that shares a knowledge base, such that knowledge accessed by said second action is not accessed by said first action.

11. The planning method of claim 6, whereby said order combination of models comprises a first model and a second model, wherein a first model state Boolean expression after a first model action is compatible with a second model state Boolean expression before execution of a second model action.

12. The planning method of claim 6, whereby the workflow creation step comprises the creation of a first workflow that comprises a first ordered combination of said models, and whereby the planning method further comprises the steps of:

creating a second workflow for solving said problem statement, said second workflow comprising a second ordered combination of said models such that said second ordered combination of models contain Boolean expressions that combine to satisfy said initial condition Boolean expression and said end condition Boolean expression,

comparing said first and said second workflows and selecting between said first and second workflows based upon a prespecified condition.

13. The method of claim 12 whereby said prespecified condition is counting numbers of actions in each of said first and said second ordered combinations and selecting the ordered combination having the lower numbers of counted actions.

14. The method of claim 12 further comprising the steps of

implementing said first workflow;

monitoring the progress of said first workflow; and

implementing said second workflow if said first workflow fails.

15. A computer program embodied on a computer readable medium for computer implemented navigation planning, the program comprising:

a first code segment for receiving a problem statement comprising an initial condition and an end condition, each of said conditions comprising one or more Boolean expressions;

a second code segments for accepting a user input to define one or more computer application models stored in computer memory, each of said models comprising a state comprising an action and one or more Boolean expressions; and

a third computer code segment for creating a workflow for solving said problem statement, said workflow comprising an ordered combination of said models such that said ordered combination of models contain Boolean expressions that combine to satisfy said initial condition Boolean expression and said end condition Boolean expression.

16. The computer program of claim 15, wherein said state defines a loop that represents a series of repeated steps according a Boolean expression associated with said loop.

17. The computer program of claim 15, wherein said state comprises plural actions which are loosely connected.

18. The computer program of claim 17, wherein said state a first action associated with said state invokes a second action that shares a knowledge base, such that knowledge accessed by said second action is not accessed by said first action.

19. The computer program of claim 15, whereby said ordered combination of models comprises a first model and a second model, wherein a first model state Boolean expression after a first model action is compatible with a second model state Boolean expression before execution of a second model action.

20. The computer program of claim 15, whereby the third computer code segment creates a first workflow that comprises a first ordered combination of said models, and whereby the computer program further comprises:

a fourth code segment for creating a second workflow for solving said problem statement, said second workflow comprising a second ordered combination of said models such that said second ordered combination of models contain Boolean expressions that combine to satisfy said initial condition Boolean expression and said end condition Boolean expression,

a fifth code segment for implementing said first workflow;

a sixth code segment for monitoring the progress of said first workflow; and

a seventh code segment for implementing said second workflow if said first workflow fails.

21. A computer program embodied on a computer readable medium for computer-implemented verification of a work flow, the program comprising:

a first code segment for receiving a problem statement comprising an initial condition and an end condition, each of said conditions comprising one or more Boolean expressions;

a second code segments for accepting a user input to define one or more computer application models stored in computer memory, each of said models comprising a state comprising an action and one or more Boolean expressions;

a third computer code segment for receiving the workflow for solving said problem statement, said workflow comprising an ordered combination of said models; and

a fourth code segment for evaluating said work flow by evaluating said ordered combination of models to determine whether the Boolean expressions associated with said ordered combination of models combine to satisfy said initial condition Boolean expression and said end condition Boolean expression.

22. The computer program of claim 21, whereby said ordered combination of models comprises a first model and a second model, and whereby the fourth code segment further assesses whether a first model state Boolean expression after a first model action is compatible with a second model state Boolean expression before execution of a second model action.