Inference method and system

Claim 2

An inference management system for use by a user with a computer system, comprising: a knowledge object having at least one expert knowledge model, and a kernel knowledge section for executing the control of, management of an inferring with different strategies using the one expert knowledge model; means for replicating at least the inferring function of the kernel knowledge section, when there are conflicting different strategies in the course of inferring by the kernel knowledge section with the at least one expert knowledge model and with at least one set of situational data, to produce plural knowledge objects (replicated), each capable of parallel inferring with another knowledge object (replicated); said knowledge objects (replicated) thereafter being means for concurrently inferring in parallel using the conflicting different strategies respectively to produce parallel inferring results, and each of the means using the at least one set of situational data used at the time of conflict; and said knowledge object including means to place its operation in a wait state during the parallel inferring of only the knowledge objects (replicated).

Claim 3

The system of claim 1, including means for removing knowledge objects (replicated).

Claim 4

The system of claim 2, including means for automatically selecting an optimum strategy from the parallel inferring results of the knowledge objects respectively using the conflicting different strategies; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting user input of a new optimum strategy different from the optimum strategy selected by the means for automatically selecting.

Claim 5

The system of claim 3, including means for automatically breaking one of the parallel inferring when a respective fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 6

The system of claim 4, wherein the corresponding knowledge object fixes said fixed period of time limit.

Claim 7

The system of claim 3, wherein said means for confirming interacts with the user through a multi-window environment.

Claim 8

The system of claim 3, including means for inputting the different strategies by the user, and wherein the kernel knowledge section includes means for storing a strategy lastly inputted by the user as the most preferred strategy for inferring together with the conflicting strategy, and means for selecting the most preferred strategy as an optimal strategy over the conflicting strategy.

Claim 9

The system of claim 1, including means for automatically selecting an optimum strategy from the parallel inferring results of the knowledge objects respectively using the conflicting different strategies; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting input of a user choice different from the optimum strategy selected by the means for automatically selecting, for adoption of a new optimum strategy.

Claim 10

The system of claim 8, including means for inputting the different strategies by the user, and wherein the kernel knowledge section includes means for storing a strategy lastly inputted by the user as the most preferred strategy for inferring together with the conflicting strategy, and means for selecting the most preferred strategy as an optimal strategy over the conflicting strategy.

Claim 11

The system of claim 1, including means for automatically breaking one of the parallel inferring when a fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 12

The system of claim 1, wherein said means for confirming interacts with the user through a multi-window environment.

Claim 13

A system of claim 1, said knowledge object including a plurality of different expert knowledge models; said kernel knowledge section selecting said one expert knowledge model from among the different expert knowledge models; and said means for replicating further replicating the at least one expert knowledge model, when there are conflicting different strategies in the course of inferring by the kernel knowledge section with the at least one expert knowledge model and with at least one set of situational data, to produce the at least one knowledge object (replicated).

Claim 14

A system of claim 1, including: said knowledge object having means for generating different hypothetical situations during inferring; said means for generating further having means for user establishing a new hypothetical situation that is not usually generated by the knowledge object during inferring when the conflicting occurs with the different strategies; and said kernel knowledge section having means for estimating the reliability and effectiveness of the different strategies by inferring with the different strategies and the new hypothetical situation to produce estimating results, and outputting the estimating results.

Claim 15

A system according to claim 13, wherein said means for estimating uses situational data in use when the conflict is generated for said estimating.

Claim 16

A system according to claim 13, including means for requesting the user to input a hypothetical situation to be analyzed for conflicting different strategies.

Claim 17

A system according to claim 13, including means for storing relationships between the conflicting different strategies and the hypothetical situations in a hypothetical case relational table; and means for fetching a hypothetical situation for a strategy to be estimated from the hypothetical case relational table.

Claim 18

A system according to claim 16, including means for storing data obtained when the conflict is generated for strategies, the different strategies, hypothetical situations used for the estimating and the estimate results with an estimate case number, and said means for fetching using the case number.

Claim 19

A system according to claim 13, including means for storing the hypothetical situations in a data table, with relationships between the different strategies and the hypothetical situations established by rules.

Claim 20

A system according to claim 13, wherein said knowledge object includes means for using a knowledge database for fetching an optimum strategy from past estimating results when the conflict is generated for different strategies and automatically selecting one of the conflicting different strategies based on the knowledge database, so that the conflicting different strategies are modified interactively.

Claim 21

An inference management method executed by a computer system, comprising: executing control of, management of, and inferring with a knowledge object comprised by at least one kernel knowledge section and a selected one of a plurality of different expert knowledge models; replicating the knowledge object by replicating at least the selected one expert knowledge model and the at least one kernel knowledge section to produce plural knowledge objects (replicated), when there is conflict in the course of inferring with situational data with use of at least one strategy and use of at least one hypothetical situation; thereafter inferring in parallel with the plural knowledge objects (replicated) that are equal in number to the number of conflicting strategies times the number of different hypothetical situations, with each duplicate kernel knowledge section respectively inferring with a unique combination of a hypothetical situation and a conflicting strategy and with a separate one of duplicate expert knowledge models, respectively, and each of the knowledge objects (replicated) using the same common situational data to produce parallel inferring results; and placing the knowledge object in a wait state during the parallel inferring of only the knowledge objects (replicated).

Claim 22

An inference management method executed by a computer system, comprising: executing control of, management of, and inferring with a knowledge object comprised by at least one kernel knowledge section and a selected one of a plurality of different expert knowledge models; replicating the knowledge object by replicating at least the selected one expert knowledge model and the at least one kernel knowledge section to produce plural knowledge objects (replicated), when there is conflict in the course of inferring with situational data with use of at least one strategy and use of at least one hypothetical situation; thereafter inferring in parallel with the plural knowledge objects (replicated) that are equal in number to the number of conflicting strategies times the number of different hypothetical situations, with each duplicate kernel knowledge section respectively inferring with a unique combination of a hypothetical situation and a conflicting strategy and with a separate one of duplicate expert knowledge models, respectively, and each of the knowledge objects (replicated) using the same common situational data to produce parallel inferring results; and placing the knowledge object in a wait state during the parallel inferring of only the knowledge objects (replicated).

Claim 23

The method of claim 20, including thereafter removing knowledge objects (replicated).

Claim 24

The method of claim 21, including automatically selecting an optimum strategy from the parallel inferring results; and confirming the selection of an optimum strategy, by outputting the inferring results for display, inquiring of the user and permitting input of a user choice among the strategies that produced the inferring results for adoption of a new optimum strategy.

Claim 25

The method of claim 22, including automatically breaking one of the parallel inferring when a fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 26

The method of claim 23, including setting the fixed period of time limit in dependence upon the respective knowledge object.

Claim 27

The method of claim 22, including outputting the parallel inferring results for display and interaction with the user through a multi-window environment.

Claim 28

The method of claim 22, including generating different strategies, and storing a strategy lastly inputted by the user as the most preferred strategy together with the strategies generated.

Claim 29

A method according to claim 20, including automatically selecting an optimum strategy by estimating the parallel inferring results; and confirming the selection of a strategy by outputting the parallel inferring results, inquiring of the user and permitting input of a user choice, among the strategies that produced the parallel inferring results, different from that selected by the automatically selecting, for adoption of a new optimum strategy.

Claim 30

A method according to claim 27, including generating different strategies and including storing a strategy inputted by the user as a preferred strategy over the strategies generated.

Claim 31

A method according to claim 20, including automatically breaking one of the parallel inferring when a fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 32

A method of claim 20, including outputting the parallel inferring results and interacting with the user through a multi-window environment.

Claim 33

The method of claim 20, including: generating different hypothetical situations and strategies for inferring with the hypothetical situation; when the conflict occurs, inputting a user established new hypothetical situation that is not usually generated for each strategy; and estimating the reliability and effectiveness of the conflicting different strategies by inferring with use of the conflicting different strategies for the user established hypothetical situation and outputting the inferring results.

Claim 34

A method according to claim 31, including using the situational data that was being used for the inferring when the conflict was generated for said estimating.

Claim 35

A method according to claim 31, including requesting the user to establish hypothetical situations to be analyzed for conflicting different strategies.

Claim 36

A method according to claim 31, including storing relationships between the conflicting different strategies and hypothetical situations in a hypothetical case relational table; and fetching a hypothetical case for a strategy to be estimated from the hypothetical case relational table at a later time with respect to a similar case.

Claim 37

A method according to claim 31, including storing data comprising user situational data, situational data in use when the conflict was generated, the different conflicting strategies, hypothetical situations used for the estimating and the estimating results with an estimate case identifier, and thereafter fetching the stored data using the case identifier for use at a later time for a similar case.

Claim 38

A method according to claim 31, including storing the hypothetical situations in a data table, with relationships between the different strategies and the hypothetical situations established by rules, for recall and use at a later time for a similar situation.

Claim 39

A method according to claim 31, including using a knowledge database for fetching an optimum strategy from past estimating results when the conflict is generated for different strategies and automatically selecting one of the conflicting strategies based on the knowledge database, so that the conflicting strategies are modified interactively.

Claim 40

An inference management system for use in a computer system, comprising: a knowledge object having an expert knowledge section with a plurality of expert knowledge models, and a kernel knowledge section for selecting, executing the control of, management of, and inferring with the selected expert knowledge model; means for generating different strategies and different hypothetical situations; means for replicating the knowledge object when there are conflicting different strategies in the course of inferring by the kernel knowledge section with the selected expert knowledge model using one set of situational data to produce plural knowledge objects (replicated) equal to the number of conflicting strategies times the number of hypothetical situations used in the inferring when the conflict occurred; means for controlling said knowledge objects (replicated) for inferring in parallel, each using a unique combination of one of the conflicting different strategies and one of the hypothetical situations respectively, and each using the same one set of situational data used at the time of conflict to produce parallel inferring results; said knowledge object including means to place its operation in a wait state during the parallel inferring of only the knowledge objects (replicated); means for thereafter removing knowledge objects (replicated); means for automatically selecting an optimum strategy from the parallel inferring results; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting input of a user choice different from the optimum strategy selected by the means for automatically selecting, for adoption of a new optimum strategy.

Claim 41

An inference management system for use in a computer system, comprising: a knowledge object having an expert knowledge section with a plurality of expert knowledge models, and a kernel knowledge section for selecting, executing the control of, management of, and inferring with the selected expert knowledge model; means for generating different strategies and different hypothetical situations; means for replicating the knowledge object when there are conflicting different strategies in the course of inferring by the kernel knowledge section with the selected expert knowledge model using one set of situational data to produce plural knowledge objects (replicated) equal to the number of conflicting strategies times the number of hypothetical situations used in the inferring when the conflict occurred; means for controlling said knowledge objects (replicated) for inferring in parallel, each using a unique combination of one of the conflicting different strategies and one of the hypothetical situations respectively, and each using the same one set of situational data used at the time of conflict to produce parallel inferring results; said knowledge object including means to place its operation in a wait state during the parallel inferring of only the knowledge objects (replicated); means for thereafter removing knowledge objects (replicated); means for automatically selecting an optimum strategy from the parallel inferring results; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting input of a user choice different from the optimum strategy selected by the means for automatically selecting, for adoption of a new optimum strategy.

Claim 42

The system of claim 38, including means for automatically breaking one of the parallel inferring when a fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 43

An inference management system for use by a user with a computer system, comprising: a knowledge object having at least one expert knowledge model, and a kernel knowledge section for executing the control of, management of and inferring with different strategies using the one expert knowledge model; said knowledge object having means for generating different hypothetical situations during inferring; means for replicating at least the inferring function of the kernel knowledge section to produce plural knowledge objects (replicated), each capable of parallel inferring with another of the knowledge objects (replicated); said knowledge objects (replicated) thereafter being means for concurrently inferring in parallel using the different hypothetical situations respectively to produce parallel inferring results; and said knowledge object including means to place its operation in a wait state during the parallel inferring of only the knowledge objects (replicated).

Claim 44

An inference management system for use by a user with a computer system, comprising: a knowledge object having at least one expert knowledge model, and a kernel knowledge section for executing the control of, management of and inferring with different strategies using the one expert knowledge model; said knowledge object having means for generating different hypothetical situations during inferring; means for replicating at least the inferring function of the kernel knowledge section to produce plural knowledge objects (replicated), each capable of parallel inferring with another of the knowledge objects (replicated); said knowledge objects (replicated) thereafter being means for concurrently inferring in parallel using the different hypothetical situations respectively to produce parallel inferring results; and said knowledge object including means to place its operation in a wait state during the parallel inferring of only the knowledge objects (replicated).

Claim 45

The system of claim 40, including means for removing knowledge objects (replicated).

Claim 46

The system of claim 41, including means for automatically selecting an optimum strategy from the parallel inferring results of the knowledge objects respectively; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting user input of a new optimum strategy different from the optimum strategy selected by the means for automatically selecting.

Claim 47

The system of claim 42, including means for automatically breaking one of the parallel inferring when a respective fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 48

The system of claim 43, wherein the corresponding knowledge object fixes said fixed period of time limit.

Claim 49

The system of claim 42, wherein said means for confirming interacts with the user through a multi-window environment.

Claim 50

The system of claim 42, including means for inputting the different strategies by the user, and wherein the kernel knowledge section includes means for storing a strategy lastly inputted by the user as the most preferred strategy for inferring, and means for selecting the most preferred strategy as an optimal strategy over the conflicting strategy.

Claim 51

The system of claim 40, including means for automatically selecting an optimum strategy from the parallel inferring results of the knowledge objects respectively using the conflicting different strategies; and means for confirming the selection of the optimum strategy by outputting the parallel inferring results, inquiring of the user and permitting input of a user choice different from the optimum strategy selected by the means for automatically selecting, for adoption of a new optimum strategy.

Claim 52

The system of claim 47, including means for inputting the different strategies by the user, and wherein the kernel knowledge section includes means for storing a strategy lastly inputted by the user as the most preferred strategy for inferring, and means for selecting the most preferred strategy as an optimal strategy over the conflicting strategy.

Claim 53

The system of claim 40, including means for automatically breaking one of the parallel inferring when a fixed period of time limit is exceeded on the one of the parallel inferring without breaking the other parallel inferring.

Claim 54

The system of claim 40, wherein said means for confirming interacts with the user through a multi-window environment.

Claim 55

A system of claim 40, wherein said knowledge object includes a plurality of different expert knowledge models; said kernel knowledge section selecting said one expect knowledge model from among the different expert knowledge models; and said means for replicating further replicating the at least one expert knowledge model to produce the at least one knowledge object (replicated).

Claim 56

A system of claim 40, including: said means for generating further including means for user establishing a new hypothetical situation that is not usually generated by the knowledge object during inferring; and said kernel knowledge section having means for estimating the reliability and effectiveness of the different strategies by inferring with the different strategies and the new hypothetical situation to produce estimating results, and outputting the estimating results.

Claim 57

A system according to claim 52 including means for requesting the user to input a hypothetical situation to be analyzed for conflicting different strategies.

Claim 58

A system according to claim 52, including means for storing relationships between the conflicting different strategies and the hypothetical situations in a hypothetical case relational table; and means for fetching a hypothetical situation for a strategy to be estimated from the hypothetical case relational table.

Claim 59

A system according to claim 54, including means for storing data obtained when the replicating occurs, the different strategies, hypothetical situations used for the estimating and the estimate results with an estimate case number, and said means for fetching using the case number.

Claim 60

A system according to claim 52, including means for storing the hypothetical situations in a data table, with relationships between the different strategies and the hypothetical situations established by rules.

Claim 61

A system according to claim 52, wherein said knowledge object includes means for using a knowledge database for fetching an optimum strategy from past estimating results when the conflict is generated for different strategies and automatically selecting one of the conflicting different strategies based on the knowledge database, so that the conflicting different strategies are modified interactively.