Knowledge based tutor

• US 4,847,784 A
• Filed: 07/13/1987
• Issued: 07/11/1989
• Est. Priority Date: 07/13/1987
• Status: Expired due to Term

• Current Assignee: Teknowledge Corporation
• Original Assignee: Teknowledge Corporation
• Inventors: Clancey, William J.
• Primary Examiner(s): Ruggiero, Joseph
• Application Number:
  US07/072,591
• Time in Patent Office: 729 Days
• Field of Search: 364/130, 364/148, 364/300, 364/513, 364/200 MS File, 364/900 MS File, 364/419, 382/14, 382/15, 434/118, 434/307, 434/308, 434/323
• US Class Current: 706/50
• CPC Class Codes: G09B 7/00 Electrically-operated teaching apparatus or devices working with questions and answers mechanically operated G09B3/00; computing arrangements G06F
• CPC Class Codes: G09B 7/04 characterised by modifying the teaching programme in response to a wrong answer, e.g. repeating the question, supplying a further explanation
• CPC Class Codes: Y10S 706/903 Control
• CPC Class Codes: Y10S 706/911 Nonmedical diagnostics
• CPC Class Codes: Y10S 706/916 Electronic or computer, internal or network, circuit
• CPC Class Codes: Y10S 706/922 Computer program preparation

Abstract

A knowledge system has a consultation system and also encodes domain-dependent tutoring knowledge as a set of conditions for interrupting the operation of the consultation system in order to evaluate a subject system. During the evaluation, the subject system is probed for its understanding of the status of the consultation system, and its understanding is compared to the actual status to obtain a measure of the subject system'"'"'s knowledge and performance relative to that of the consultation system. The direction of the probing and the source of information for instruction or diagnosis is based upon the condition causing the interruption of the consultation system. Preferably an authoring system scans the domain-dependent knowledge base of the consultation system and determines a set of possible interrupt conditions. From this set a user exercises judgment in selecting a subset of conditions that are appropriate for the subject domain and the needs of the student. The selected conditions and a selected test case dialog are stored in a case file, and a number of different case files may be stored in a case library. The authoring system preferably creates a file or index of tutorial knowledge which correlates the relevant domain knowledge with the interrupt conditions. The tutorial knowledge includes, for example, expressions for causing interrupts after their values are found, rules concluding the expressions, the values concluded by the rules, and the factors in the rules.

Claims

1. A knowledge system comprising a computer having a memory, said memory storing a knowledge base encoding knowledge about a particular subject domain, said computer having

a knowledge base interpreter for interpreting said knowledge base to solve specified problems pertaining to said subject domain and to indicate the methods of solution to a subject system, said memory further storing information about the status of the solution of said problems by said knowledge base interpreter, said knowledge base interpreter including means for changing said information about the status of the solution of said problems during said interpreting of said knowledge base, and

an evaluation system communicating with said subject system and accessing the knowledge in said knowledge base, said evaluation system including means for probing said subject system about selected portions of said knowledge about said subject domain to obtain responses from said subject system, and means for comparing said responses from said subject system to said selected portions of said knowledge about said subject domain to evaluate the subject system'"'"'s knowledge and performance with respect to said subject domain, said means for probing and said means for comparing being operative during the operation of said knowledge base interpreter,

wherein said knowledge base further encodes knowledge defining a set of interrupt conditions of said status of said solution of said problems, and

wherein said computer further includes means for determining whether any of said interrupt conditions occur in said status of the solution of said problems and thereupon interrupting the interpreting of said knowledge base to solve said problems and activating said evaluating system to probe said subject system about the interrupt condition having occurred so that the subject system'"'"'s knowledge and performance are evaluated while the methods of solution to said specified problems are indicated to said subject system.

2. The knowledge system as claimed in claim 1, wherein said means for comparing includes means for interpreting the knowledge in said knowledge base to determine the knowledge about said subject domain that is relevant to said conditions.

3. The knowledge system as claimed in claim 1, wherein said means for probing includes means for successively probing about said interrupt condition having occurred in response to said means for comparing in order to tailor said probing to the performance of said subject system.

4. The knowledge system as claimed in claim 3, wherein said memory further stores a secondary knowledge base indicating how to tailor said probing to the performance of said subject system, and said means for successively probing includes means for interpreting said secondary knowledge base in order to tailor said probing to the performance of said subject system.

5. The knowledge system as claimed in claim 1, wherein said memory stores a number of case files including knowledge defining a plurality of problems and respective predefined interrupt conditions for said problems, said knowledge system includes means for selecting a desired one of said case files for interpreting said knowledge base to solve a respective one of said problems by reading from said memory said information defining the selected problem, and said means for probing probes said subject system about the respective interrupt conditions for the selected problem.

6. The knowledge system as claimed in claim 1, wherein said means for probing includes means for probing said subject system for conclusions, means for probing said subject system for factors that support said conclusions, and means for probing said subject system for values of factors that are consistent with said conclusions.

7. The knowledge system as claimed in claim 6, wherein said means for comparing includes means for comparing factor identified by said subject system to said factors that are consistent with said conclusions to determine missed factors which are consistent with said conclusions but not identified by said subject system, and said means for probing includes means for probing the subject system for values of said missed factors.

8. The knowledge system as claimed in claim 7, wherein said means for comparing includes means for comparing values identified by said subject for said missed factors to actual values used by said knowledge system for said missed factors, and means for indicating portions of said knowledge base which mention the actual values of missed factors when the values identified by said subject system differ from the actual values of said missed factors.

9. The knowledge system as claimed in claim 1, wherein said means for comparing determines faults of said subject system, and said evaluation system includes means for responding to some of said faults and ignoring others of said faults.

10. The knowledge system as claimed in claim 1, wherein said evaluation system is independent of said subject domain.

11. The knowledge system as claimed in claim 1, wherein said evaluation system further comprises means for generating a record of the subject system'"'"'s knowledge and performance.

12. The knowledge system as claimed in claim 1, wherein said memory stores an index indicating the location of certain information stored in said knowledge base about said subject domain, and wherein said evaluation system includes means for accessing said knowledge base by reading said index to determine the location of said certain information, and then reading said knowledge base at the determined location.

13. The knowledge system as claimed in claim 12, wherein said index indicates the locations of rules pertaining to said interrupt conditions.

14. The knowledge system as claimed in claim 13, wherein said index also indicates the locations of factors in the premises of said rules pertaining to said interrupt conditions.

15. The knowledge system as claimed in claim 1, wherein said evaluation system includes a control procedure stored in said memory, written in the same language as said knowledge base encoding knowledge about the particular subject domain, and interpreted by said knowledge base interpreter.

16. A knowledge system comprising a computer having a memory, said memory storing a knowledge base including rules of inference encoding knowledge about a particular subject domain, said computer including an interference engine for executing an inference procedure to apply said rules of inference to solve specified problems pertaining to said subject domain, wherein

said computer includes an evaluation system for accessing said knowledge base and evaluating the status of the application of said rules of inference in the context of a specified one of a set of predetermined conditions, said memory stores said set of predetermined conditions for specifying when said evaluation system is to evaluate said status, and said inference procedure includes a step for testing whether said predetermined conditions currently hold for said status, and when one of said conditions is found to hold, thereupon interrupting the operation of said inference engine, and initiating operation of said evaluation system to evaluate said status in the context of the condition found to hold.

17. The knowledge system as claimed in claim 16, wherein said set of predetermined conditions is specified by a predefined set of expressions which have values concluded by said rules, said step of said inference procedure is the step at which the value of an expression is found, and said conditions are found to hold when the expression having its value found is a member of said predefined set of expressions.

18. The knowledge system as claimed in claim 17, wherein said evaluation system includes means for determining which rules concluded about the expression having its value found, and means for determining which factors in the rules caused said rules to conclude about the expression having its values found.

19. The knowledge system as claimed in claim 16, wherein said evaluation system includes means for probing a subject system about said predetermined conditions to obtain responses from said subject system, and means for comparing said responses to knowledge in said knowledge base pertaining to said conditions.

20. A knowledge system comprising a computer having a memory, said memory storing a knowledge base capable of being interpreted by a knowledge base interpreter to solve specified problems pertaining to a subject domain, said computer further comprising an authoring system for designing an evaluation procedure, said authoring system comprising means for analyzing the content of said knowledge base to determine a set of interrupt conditions for interrupting said interpreting of said knowledge base when said conditions occur to thereupon initiate said evaluation procedure, and means for designing said evaluation procedure for comparing knowledge about a subset of said interrupt conditions to knowledge obtained from a subject system.

21. The knowledge system as claimed in claim 20, wherein said means for designing said evaluation procedure includes means for presenting to a human user an indication of said set of interrupt conditions, and receiving from said human user an indication of said subset of said interrupt conditions.

22. The knowledge system as claimed in claim 20, wherein said authoring system further comprises means for storing in a secondary knowledge base information defining the interrupt conditions in said subset.

23. The knowledge system as claimed in claim 20, wherein said authoring system is independent of said subject domain.

24. The knowledge system as claimed in claim 20, wherein said knowledge base includes rules which conclude values for expressions, and said means for analyzing includes means for determining which of said rules cannot be discussed by said evaluation procedure.

25. The knowledge system as claimed in claim 20, wherein said knowledge base defines expressions which take on values when said knowledge base is interpreted by said knowledge base interpreter, and said means for analyzing includes means for determining which of said expressions cannot be used as said interrupt conditions.

26. The knowledge system as claimed in claim 20, wherein said knowledge base includes rules and facts which conclude values for expressions, and wherein said means for analyzing includes means for integrating said rules and facts by combining values concluded for an expression by rules with values concluded for the expression by facts.

27. The knowledge system as claimed in claim 20, wherein said knowledge base includes rules which conclude values for expressions, said rules including rules containing variables, and said means for analyzing includes means for determining conclusions a rule with variables can make, and means for instantiating said variables to determine expressions that can be included in the premises of said rules.

28. The knowledge system as claimed in claim 20, wherein said knowledge base includes rules which conclude values for expressions, said subset of said interrupt conditions are indicated by a list of selected ones of said expressions, and said means for analyzing includes means for finding the rules that conclude the selected expressions, means for finding the values concluded by the rules that conclude the selected expressions, means for finding the factors in the premises of the rules that conclude the selected expressions, and means for finding the values concluded about the selected expressions by the rules.

29. The knowledge system as claimed in claim 28, wherein said interrupt conditions appear in said knowledge base and said authoring system further comprises means for storing in said memory pointers to said interrupt conditions appearing in said knowledge base.

30. The authoring system as claimed in claim 20, wherein said authoring system is implemented as a procedure written in the same language as said knowledge base and executed by said knowledge base interpreter.

31. A knowledge system comprising a computer having a memory, said memory storing a knowledge base capable of being interpreted by a knowledge base interpreter to solve specified problems pertaining to a subject domain, said computer comprising an authoring system for designing a domain dependent evaluation procedure, said authoring system includign means for analyzing the content of said knowledge base to determine interrupt conditions defined at locations in said knowledge base for interrupting said interpreting of said knowledge base when said conditions occur to thereupon initiate said evaluation procedure, and means for generating an index of said conditions, said index including pointers to the locations of said conditions in said knowledge base.

32. The knowledge system as claimed in claim 31, wherein said means for generating includes means for adding new entries to said knowledge base.

34. The knowledge system as claimed in claim 31, further including means for deleting portions of said knowledge base.

35. The knowledge system as claimed in claim 31, wherein said knowledge base includes entries, said knowledge base stores information explaining why respective entries were made, and wherein said means for analyzing includes means for accessing the information explaining why the respective entries were made.

33. The knowledge system as claimed in 32, wherein said means for analyzing includes means for analyzing entries added to the knowledge base by said means for adding.

36. A knowledge system comprising a computer having a memory, said memory storing a knowledge base encoding knowledge about a particular subject domain, said computer having a knowledge base interpreter for interpreting said knowledge base to solve specified problems pertaining to said subject domain and indicating the methods of solution to a subject system,

wherein said computer further comprises an authoring system and an evaluation system, said authoring system including:

means for analyzing said knowledge base to identify conditions of interest suitable for interrupting the interpreting of said knowledge base while solving specified problems pertaining to said subject domain,

means for presenting said conditions of interest to a human user for selection of a subset of said conditions suitable for comparing the performance of a subject system to the performance of said knowledge base interpreter for solving said problems, and

means for receiving an indication of said subset of said conditions, and

said evaluation system including;

means for interrupting the operation of said knowledge base interpreter upon the occurrence of a condition included in said subset of said conditions, and

means operative upon said interrupting for probing said subject system to obtain a response about the status of the solution of said problem by said knowledge base interpreter, and

means for comparing said response to the actual status of the solution of said problem by said knowledge base interpreter.

37. The knowledge system as claimed in claim 36, wherein said authoring system and said evaluation system are domain-independent.

38. The knowledge system as claimed in claim 36, wherein said authoring system and said evaluation system are implemented as procedures interpreted by said knowledge base interpreter and written in the same language as said knowledge base encoding knowledge about the subject domain.

39. The knowledge system as claimed in claim 36, wherein said knowledge base includes rules having factors in premises and concluding values for expressions, and said means for probing includes:

means for probing the subject system for a value of an expression pertaining to said condition upon which the operation of the knowledge base interpreter is interrupted.

means for probing the subject system for factors supporting the value of said expression pertaining to said condition, and

means for probing the subject system for the values of factors supporting the values of said expression pertaining to said condition.

40. The knowledge system as claimed in claim 39, wherein said subset of conditions is indicated by a selected subset of said expressions, and said conditions occur when values are found for the expressions included in said subset of expressions.

41. A method of comparing the performance of a subject system to the performance of a knowledge system, said knowledge system comprising a computer having a memory, said memory storing a knowledge base encoding knowledge about a particular subject domain, said computer having a knowledge base interpreter for interpreting said knowledge base to solve specified problems pertaining to said subject domain and indicating the methods of solution to a subject system, said method comprising the steps of:

(a) operating said computer to scan the knowledge base and determine a set of conditions of interest suitable for interrupting the interpretation of said knowledge base while solving specified problems pertaining to said subject domain,

(b) operating said computer to transmit to a human user an indication of said set of conditions of interest,

(c) operating said computer to receive from said human user an indication of a subset of said set of conditions,

(d) operating said knowledge base interpreter to solve a specified problem pertaining to said subject domain, and while operating said knowledge base interpreter, operating said computer to test for the occurrence of said conditions included in said subset of conditions, and upon the occurrence of a condition included in said subset of conditions, to interrupt the operation of said knowledge base interpreter and to

(e) probe the subject system to obtain a response about the status of the solution of said problem by said knowledge base interpreter, and

(f) compare said response to the actual status of the solution of said problem by said knowledge base interpreter.

42. The method as claimed in claim 41, wherein said knowledge base includes rules concluding values for expressions, and said step (a) of operating said computer to scan the knowledge base includes determining a list of expressions concluded by the rules, said step (b) includes transmitting to the human user said list of expressions concluded by the rules, and step (c) includes receiving from the user a subset of the expressions included in said list, said step (d) includes interrupting the operation of the knowledge base interpreter each time when the knowledge base interpreter finds the value for an expression in said subset, said step (e) probes the subject system for information pertaining to the expression which had its value found and caused the interruption of said knowledge base interpreter.

43. The method as claimed in claim 42, wherein said rules have premises including factors which taken on values, and wherein said step (e) includes operating said computer to probe the subject system for the values of factors in rules which conclude values for the expression which had its value found and caused the interruption of said knowledge base interpreter.

44. The method as claimed in claim 41, further comprising the steps of:

(g) recording in memory a file of test cases including data defining specific problems pertaining to said subject domain and subsets of said conditions of interest, said data defining the specific cases being suitable for input to said knowledge base interpreter; and

(h) performing steps (c) to (f) using a selected one of said test cases to specify said specified problem and to specify the subset of conditions for interrupting the operation of the knowledge base interpreter.

45. The method as claimed in claim 41, further comprising the steps of operating said computer to scan the knowledge base for information pertaining to the conditions, operating said computer to store in said memory an index of said information pertaining to each of said conditions in said subset of conditions, and after interruption of said knowledge base interpreter, operating said computer to read said index of information pertaining to the respective condition causing interruption of the knowledge base interpreter in order to find information about the status of the solution of said problem by said knowledge base interpreter.