System for automatically and transparently mapping rules and objects from a stable storage database management system within a forward chaining or backward chaining inference cycle

• US 5,136,523 A
• Filed: 03/26/1991
• Issued: 08/04/1992
• Est. Priority Date: 06/30/1988
• Status: Expired due to Term

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Digital Equipment Corporation (HP Inc.)
• Inventors: Landers, Terry A.
• Primary Examiner(s): MacDonald, Allen R.
• Application Number:
  US07/679,155
• Time in Patent Office: 497 Days
• Field of Search: 364/513, 364/200, 364/900
• US Class Current: 706/50
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

A system for mapping rules and objects from a database system during an inference cycle includes a database system a knowledge request translator, and an inference system. The database system stores both the facts and rules. The knowledge request translator, in response to rule retrieval requests, retrieves rules from the database system, each rule retrieval being based on rule selection criteria that are included in the rule request. The inference system generates rule retrieval requests, including rule selection criteria, for use by the knowledge request translator. The inference system initiates inference operations in connection with rules retrieved by said knowledge request translator, and facts in the database system. In initiating inference operations, the inference system supplies fact criteria to the knowledge request translator, which generates, in response thereto, database inquiries to initiate pattern matching operations in connection with the facts in the database system. The results of the pattern matching operations are used in connection with subsequent processing during the inference operations.

Claims

1. A system comprising:

A. a database system including a plurality of records for storing facts and rules, said database system being accessible by a plurality of operators to allow each one of said operators to cause said facts and rules to be retrieved from said database system, said database system including means for managing access to said database system to resolve conflicting requests by said operators for access to said rules;

B. a knowledge request translator responsive to rule retrieval requests for generating database requests to enable the retrieval of rules from said database system, each rule retrieval being based on rule selection criteria included in said rule retrieval request; and

C. an inference system, responsive to commands from each one of said operators, for generating rule retrieval requests, including rule selection criteria, and for performing inference operations in connection with rules retrieved from said database system by said knowledge request translator, said inference system enabling the retrieval of facts from said database system for use in connection with the inference operations.

2. A system as defined in claim 1 in which said inference system includes a rule storage area for storing rules and an inference engine, said inference engine generating rule retrieval requests and performing inference operations in connection with rules stored in said rule storage area, said knowledge request translator storing rules retrieved in response to rule retrieval requests in said rule storage area.

3. A system as defined in claim 1 in which said inference system includes:

A. a rule retrieval request portion that generates rule retrieval requests, each including rule selection criteria; and

B. a rule processing portion that performs inference operations in connection with the rules retrieved by the knowledge request translator in response to the rule retrieval request.

4. A system as defined in claim 3, each said rule including a rule value expression element identifying at least one relational operation in connection with a fact in said database system and a value expression in said rule, in which said rule processing portion includes:

A. a constant processing portion for processing rule value expression elements whose value expressions are constant values in connection with facts in said database system to identify a logical conclusion;

B. a variable processing portion for processing rule value expression elements whose value expressions are variable values in connection with facts in said database system to generate a binding list to identify the values of said variable value expressions which provide a selected logical conclusion; and

C. a control portion for controlling the operation of said constant processing portion and said variable processing portion.

5. A system as defined in claim 4 in which said control portion first enables said constant processing portion to process rule value expression elements of rules all of whose rule value expression elements are constant values and subsequently enables said variable processing portion and said constant processing portion to process rule value expression elements of rules at least one of whose rule value expression elements is a variable value.

6. A system as defined in claim 4 in which:

A. the constant processing portion generates a constant pattern-matching request in response to a rule value expression element;

B. the knowledge request translator generates a database inquiry in response to the constant pattern-matching request, the database inquiry identifying the rule value expression element; and

C. the database system performs an existence operation in connection with the rule value expression element.

7. A system as defined in claim 6 in which a rule value expression element identifies a relation by class name and a value, the database system performing the existence operation by determining whether a relation exists which is identified by the class name and which contains the value.

8. A system as defined by claim 7 in which the rule value expression element further identifies a slot name associated with the value, the database system further performing the existence operation by determining whether a relation exists which is identified by the class name and slot name and which contains the value.

9. A system as defined in claim 4 in which:

A. the variable processing portion generates a variable pattern-matching request in response to a rule value expression element;

B. the knowledge request translator generates a database inquiry in response to the variable pattern-matching request, the database inquiry identifying the rule value expression element; and

C. the database system performs a binding list generation operation in connection with the rule value expression element to identify variable values.

10. A system as defined in claim 9 in which a rule value expression element identifies a relation by class name and a value, the database system performing the binding list generation operation by identifying selected values associated with a relation which is identified by the class name.

11. A system as defined by claim 10 in which the rule value expression element further identifies a slot name associated with the value, the database system further performing the binding list generation operation by identifying selected values associated with a relation which is identified by the class name and slot name.

12. A system as defined in claim 1 in which the inference system operates in response to commands specified by said rules.

13. A system as defined in claim 12 in which the rules comprise rule elements at least some of which identify a relation identified by class name and include an update command that enables a relation, identified by a class name, to be updated, the inference system generating a rule retrieval request identifying the class name enabling the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name.

14. A system as defined in claim 13 in which the database system includes rules comprising a plurality of records each including a class name field, the database system, in response to a database request, retrieves records the contents of whose class name field corresponds to the class name in the rule retrieval request.

15. A system as defined in claim 13 in which at least some of the rule elements further identify a slot in the identified relation, the slot being further identified by a slot name, the rule retrieval request generated by the inference system further identifying the slot name which enables the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name and the slot name.

16. A system as defined in claim 15 in which the database system includes rules comprising a plurality of records each including a class name field and at least some including a slot name field, the database system, in response to a database request, retrieving records the contents of whose class name field corresponds to the class name in the rule retrieval request and the slot name in the rule retrieval request.

17. A system as defined in claim 12 in which the rules comprise rule elements at least some of which identify a relation identified by class name and include a create command that enables a relation, identified by a class name, to be created, the inference system generating a rule retrieval request identifying the class name enabling the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name.

18. A system as defined in claim 17 in which the database system includes rules comprising a plurality of records each including a class name field, the database system, in response to a database request, retrieves records the contents of whose class name field corresponds to the class name in the rule retrieval request.

19. A system as defined in claim 17 in which at least some of the rule elements further identify a slot in the identified relation, the slot being further identified by a slot name, the rule retrieval request generated by the inference system further identifying the slot name which enables the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name and the slot name.

20. A system as defined in claim 19 in which the database system includes rules comprising a plurality of records each including a class name field and at least some including a slot name field, the database system, in response to a database request, retrieving records the contents of whose class name field corresponds to the class name in the rule retrieval request and the slot name in the rule retrieval request.

21. A system as defined in claim 12 in which the rules comprise rule elements at least some of which identify a relation identified by class name and include a delete command that enables a relation, identified by a class name, to be deleted, the inference system generating a rule retrieval request identifying the class name enabling the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name.

22. A system as defined in claim 21 in which the database system includes rules comprising a plurality of records each including a class name field and a negated flag, the database system, in response to a database request, retrieves records the contents of whose class name field corresponds to the class name in the rule retrieval request and whose negated flags are set.

23. A system as defined in claim 21 in which at least some of the rule elements further identify a slot in the identified relation, the slot being further identified by a slot name, the rule retrieval request generated by the inference system further identifying the slot name which enables the knowledge request translator to, in turn, generate a database request to enable retrieval of rules from the database system associated with the class name and the slot name.

24. A system as defined in claim 23 in which the database system includes rules comprising a plurality of records each including a class name field and at least some including a slot name field, the database system, in response to a database request, retrieving records the contents of whose class name field corresponds to the class name in the rule retrieval support and the slot name in the rule retrieval request.

25. A system as defined in claim 12 in which the rules including backward-chaining rules that comprise rule elements at least some of which identify a relation identified by class name, and include a backward chaining command, the inference system generating a rule retrieval request identifying the class name enabling the knowledge request translator to, in turn, generate a database request to enable retrieval of backward-chaining rules from the database system associated with the class name.

26. A system as defined in claim 25 in which the database system includes backward-chaining rules comprising a plurality of records each including a class name field, the database system, in response to a database request, retrieves records the contents of whose class name field corresponds to the class name in the rule retrieval request.

27. A system as defined in claim 25 in which at least some of the rule elements further identify a slot in the identified relation, the slot being further identified by a slot name, the rule retrieval request generated by the inference system further identifying the slot name which enables the knowledge request translator to, in turn, generate a database request to enable retrieval of backward-chaining rules from the database system associated with the class name and the slot name.

28. A system as defined in claim 27 in which the database system includes backward-chaining rules comprising a plurality of records each including a class name field and at least some including a slot name field, the database system, in response to a database request, retrieving records the contents of whose class name field corresponds to the class name in the rule retrieval request and the slot name in the rule retrieval request.

29. A system as defined in claim 1, each rule including at least one rule value expression element having an associated name, relational operation and value expression, each rule retrieval request including, as the selection criteria, a name, said knowledge request translator enabling the retrieval from said database system of rules having at least one rule value expression element which includes a name corresponding to the name included with the rule retrieval request.

30. A system as defined in claim 29, the name associated with each rule value expression element having a major portion and a minor portion, each rule retrieval request including, as the selection criteria, a name having a major portion and a minor portion, said knowledge request translator enabling a rule retrieval using both the major portion and minor portion of said name.

31. A system as defined in claim 29 in which said database system includes:

A. a rule storage portion including a plurality of records each for storing a rule value expression element, said record including a name field for storing a name, and

B. a database management portion for retrieving the contents of said rule storage portion in response to database requests from said knowledge request translator each including a name, the database management portion retrieving records from said rule storage portion the contents of whose name fields correspond to the name in said database requests.

32. A system as defined in claim 31 in which each record further includes a rule identification field for storing the identification of a rule, said knowledge request translator enabling the database management portion to retrieve, in response to a database request, all of the records relating to the rule identified by the contents of the record including the name retrieved in response to the database request.

33. A system as defined in claim 32, said record further including a sequence identification field identifying the sequence of said rule value expression element in said rule, said knowledge request translator including a rule assembly portion for assembling a rule in the order identified in the sequence identification field in response to the receipt of the records relating to the rule, the assembled rule to thereafter be used by the inference system in connection with a inference operation.

34. A system comprising:

A. a database system including a plurality of records for storing facts and rules, said database system being accessible by a plurality of operators to allow each one of said operators to cause said facts and rules to be retrieved from said database system, said database system including means for managing access to said database system to resolve conflicting requests by said operators for access to said rules;

B. a knowledge request translator responsive to rule retrieval requests for generating database requests to enable the retrieval of rules from said database system, each rule retrieval being based on rule selection criteria included in said rule retrieval request, said knowledge request translator assembling the retrieved rules in a rule virtual memory; and

C. an inference system, responsive to commands from each one of said operators, for generating rule retrieval requests, including rule selection criteria, and for performing inference operations in connection with rules in said rule virtual memory, said inference system enabling the retrieval of facts from said database system for use in connection with the inference operations.

35. A method as defined in claim 34, each rule including at least one rule value expression element having an associated name, relational operation and value expression, each rule retrieval request including, as the selection criteria, a name, rule retrieval step including the step of determining whether a rule includes at least one rule value expression element which includes a name corresponding to the name included with the rule retrieval request.

36. A method as defined in claim 35, the name associated with each rule value expression element having a major portion and a minor portion and each rule retrieval request including, as the selection criteria, a name having a major portion and a minor portion, the rule retrieval step including the steps of determining whether a rule includes at least one rule value expression element which includes a major portion which corresponds to the major portion of said selection criteria and said rule value expression element which includes a minor portion which corresponds to the minor portion of said selection criteria.

37. A method as defined in claim 35 in which said database system includes (i) a rule storage portion including a plurality of records each for storing a rule value expression element, said record including a name field for storing a name and a rule identification field for storing the identification of a rule, and (ii) a database management portion for retrieving the contents of said rule storage portion in response to database requests from said knowledge request translator each including a name, the database management portion retrieving records from said rule storage portion the contents of whose name fields correspond to the name in said database requests, said rule retrieval step including the step of retrieving, in response to a database request, all of the records relating to the rule identified by the contents of the record including the name retrieved in response to the database request.

38. A method as defined in claim 37 said record further including a sequence identification field identifying the sequence of said rule value expression element in said rule, said rule retrieval step including the step of assembling a rule, using the contents of the sequence identification field of the retrieved records in response to the receipt of the records relating to the rule.

39. A method comprising the steps of:

A. maintaining a database system including a plurality of records for storing facts and rules;

B. permitting each one of a plurality of operators to access said database system to cause said facts and rules to be retrieved therefrom and managing access to said database system to resolve conflicting requests by said operators for access to said rules;

C. generating a rule retrieval request including rule selection criteria in response to commands from each one of said operators;

D. retrieving rules from said database system in response to the rule selection criteria and assembling the retrieved rules in a rule virtual memory; and

E. performing an inference operation in connection with rules in said rule virtual memory in connection with facts from said database system.

40. A method as defined in claim 39, each said rule including a rule value expression element identifying at least one relational operation in connection with a fact in said database system and a value expression in said rule, in which the inference operation performance step includes the steps of:

A. processing rule value expression elements of rules all of whose rule value expression elements include value expressions which are constant values in connection with facts in said database system to identify a logical conclusion;

B. processing rule value expression elements which contain value expressions that include variable values, of rules which include rule value expression elements whose value expressions include variable values, in connection with facts in said database system to generate a binding list to identify the values of said variable value expressions which provide a selected logical conclusion; and

C. processing rule value expression elements which contain value expressions that include constant values, of rules which include rule value expression elements whose value expressions include variable values, in connection with facts in said database system to provide the selected logical conclusion.