Method for representation of knowledge in a computer as a network database system

• US 5,379,366 A
• Filed: 01/29/1993
• Issued: 01/03/1995
• Est. Priority Date: 01/29/1993
• Status: Expired due to Term

• Current Assignee: Reachable, Inc.
• Original Assignee: Dallas B. Noyes
• Inventors: Noyes, Dallas B.
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): Dorvil, Richemond
• Application Number:
  US08/011,355
• Time in Patent Office: 704 Days
• Field of Search: 395/51, 395/11, 395/12, 395/275, 395/600, 395/54, 395/60
• US Class Current: 706/55
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition

Abstract

A system for knowledge representation in a computer, together with the ability to recognize, store and use patterns in the knowledge representation, together with the ability for Natural Language Interaction with the knowledge representation system, together with systems to automatically transform information in the knowledge representation into a multitude of documents or other human interpretable displays in a plurality of different formats or views. User interaction with the knowledge representation through the view documents is achievable through a multitude of various possible formats. The Knowledge Representation system defines a novel database engine constituting a method for modeling knowledge as a network of concepts and a plurality of relationships between the concepts comprising the network. Each concept is represented as a record in the database which is identified by a unique record reference number. The unique record reference numbers are stored within the records comprising the database to record the plurality of relationships between concepts.

Claims

1. A method for representing information in a computer system, comprising the steps of:

establishing in said computer system a knowledge representation database made up of individual records, wherein each record is associated with a unique reference number (URN) which identifies each record and wherein each record stores at least one relationship comprised of a characterization and a value,

the characterization of said relationship being a URN of a second record which defines [the]a nature of said relationship, and

the value of said relationship being a complex data representation composed of at least one internal value, external value, or mixed value which define an object of said relationship,

internal values storing only URNs of other records,

external values storing external data such as character strings, integers, and real numbers, and

mixed values storing a combination of internal and external values;

establishing an index to said knowledge representation database made up of the name of each record together with the associated URN of said record, wherein the name of the record is an external value of a relationship stored therein which designates that external value as a character string description of a concept represented by said record;

establishing, for each record in said knowledge representation database, fundamental relationships between said record and other records in said database, said fundamental relationships being comprised of

intrastratum relationships which store URNs of other records on the same strata or level of abstraction designated as separate libraries within the knowledge representation system, said intrastratum relationships being designated as parent and children relationships which identify the record in which said parent and children relationships are stored as a member of the same library as the records identified by the URNs stored in said intrastratum relationships, and

interstrata relationships which store URNs of other records in different strata or libraries, said interstrata relationships being designated as Type record relationships which identify the record in which said Type record relationships are stored as a particular instance of records in another stratum or library;

designating certain records as system concepts by storing the URNs of said certain records in a system concept index to said database reserved for records which represent system concepts of said knowledge representation database,

system concept records being records which are used as termination points of networks of said fundamental relationships and which are recognized by the system by determining whether the URN or the name of a particular record is in said system concept index, wherein

system concept records designating strata or libraries (such as System library, Attribute library, Component library, and Project library) are the termination of parent fundamental relationships,

system concept records designating attribute classes (such as Assignment, Connection, Non-Binding, Rules, and External) are the termination of parent fundamental relationships for records which are descendants thereof and which store the URN of the Attribute library system concept as a parent relationship, and

system concept records designating attribute properties (such as Name, Data Type, Field Length, and Prompt) are the termination of relationship characterization networks,

said system concepts being required to store only fundamental relationships;

storing within each record comprising said Attribute library at least one relationship which is characterized by a URN of an attribute property system concept record, wherein said at least one relationship stores the value of the name of the concept represented by the record in which said attribute property System record URN is stored;

storing within each record comprising said Component library at least one relationship which is characterized by a URN of an Attribute library record, wherein said at least one relationship stores the value of the name of the concept represented by the record in which said Attribute library record URN is stored;

storing within each record comprising said Project library at least one relationship which is characterized by a URN of a Component library record, wherein said at least one relationship stores the value of the name of the concept represented by the record in which said Component library record URN is stored;

establishing in said computer system at least one editor for modifying the records and relationships stored in said database, including means for recognizing patterns in the relationships stored in said records;

storing said recognized patterns as relationships in the records associated with the recognized patterns;

establishing an additional class of system concept records to identify relationships storing values that define said recognized patterns, wherein each of said additional class system concept records represent particular types of patterns in said relationships;

operating on said stored patterns in the operation of said at least one editor by reading relationships storing patterns predetermined to be relevant to said at least one editor and using the values of said relationships in limiting the operation of said editor, said relationships storing patterns relevant to said editor being identified by the characterization of said relationships as system concepts identified by the system as being relevant to said editor;

establishing in said computer a descriptive database for describing an active concept record designated by a user, comprised of a plurality of records each of which stores a single relationship having an associated URN for said active concept record, and an associated URN for a source record in which said relationship is stored,

the URN for said active concept being the URN of the record in said knowledge representation database for which the description in the descriptive database is assembled, and

the URN for said source record being the URN of that record in said knowledge representation database in which said relationship is stored;

reading a descriptive network for said active concept by reading all records in said knowledge representation database forming a network of related records through the fundamental relationships of parent and type, combining the relationship lists from said read records, and storing said relationships from said read records in said descriptive database,

said relationship lists being combined by applying Taxonomy, Type, Composition and User inheritance rules,

said relationships being stored in said descriptive database together with the URN of said active concept and the URN of said source record;

selecting a view, class and type of display for said active concept and deriving the selected type of display by assigning icons to the records representing concepts in said descriptive database according to the type of display selected, organizing and locating said icons in a display space of said computer according to the selected class, and creating connection icons for interconnections between concept icons located in said display space according to the selected view; and

interacting with said knowledge representation database through interaction with said icons in said display space and evaluation of said icon interaction through the use of decision trees for evaluation of the view, command history, system flags, and icon association for identification of appropriate responses to said icon interaction.

2. The method for representing information in a computer system according to claim 1, further comprising the step of establishing at least one optional user relationship in the knowledge representation database by specifying a user defined Attribute other than a system concept, and storing said user defined Attribute as a record in said knowledge representation database, wherein the URN of said user defined Attribute is stored in records having said user defined Attribute as the characterization of a relationship stored therein.

3. The method for representing information in a computer system according to claim 1, further comprising the step of establishing in said computer system a plurality of editors for modifying the records of said database and relationships stored in said records, wherein said editors enable users to create, read and replace information in each of the records in the knowledge representation database, and enable users to identify and store relationships in the records in the knowledge representation database by specifying the characterization and value of a relationship and storing the specified relationship in a record.

4. The method for representing information in a computer system according to claim 1, wherein said patterns in said relationships are comprised of recurring correlations in values of fundamental relationships in Project library records, and recurring correlations in values of user relationships in records of the Project and Component libraries.

5. The method for representing information in a computer system according to claim 4, further comprising the step of storing patterns recognized in Project library records in the component record represented by the URN stored in the type fundamental relationship of said Project records or in a parent record of said component record, and

storing patterns recognized in Component library records in the component record in which they are first identified or in a parent record of said component record;

the characterization of the relationship storing the recognized pattern being designated as a system concept.

6. The method for representing information in a computer system according to claim 1, further comprising the steps of:

evaluating a natural language user input expression by sequential application of an augmented transition network parse based on identification of the function of a recognized expression in said knowledge representation database, and by parsing by differences based on the evaluation of said recognized expression and said identified function;

wherein a recognized expression corresponds to a name associated with a record in the knowledge representation database, and the function of a recognized expression is evaluated by identifying the system concepts in a descriptive network of recognized expressions.

7. The method for representing information in a computer system according to claim 6, further comprising the steps of dynamically defining concepts not found in said knowledge representation database and adding said defined concepts to the knowledge representation database in a particular location within a particular library through evaluation of said input expression as a constraint on the fundamental relationships of said dynamically defined concepts.

8. The method for representing information in a computer system according to claim 6, further comprising the steps of coordinating the derivation of appropriate views of the information in the knowledge representation database in response to said natural language user input expression through evaluation of the input expression for specifying the active concepts and selected appropriate view, class and type of display for response to the input expression.

9. A method for representing information in a computer system, comprising the steps of:

establishing in said computer system a knowledge representation database made up of individual records, wherein each record is associated with a unique reference number (URN) which identifies each record;

storing in each of a plurality of records of said knowledge representation database the URN of at least one other record of said knowledge representation database; and

establishing means in said computer system for automatically deriving a human understandable format for displaying information contained in said records through reading and evaluation of said stored URNs.

10. The method for representing information in a computer system according to claim 9, further comprising the step of storing in at least one record the URN of a second record with said second record storing the URN of said at least one record.

11. The method for representing information in a computer system according to claim 9, wherein each record of said database stores a plurality of relationships, said relationships being comprised of a characterization and a value,

the characterization of said relationship being a URN of a second record which defines the nature of the relationship, and

the value of said relationship being a complex data type comprised of internal, external and mixed values which define the object of a relationship,

internal values storing only URNs of other records and,

external values storing values other than the URNs of other records such as character strings, integers, and real numbers, and

mixed values storing a combination of values typical of internal and external values.

12. The method for representing information in a computer system according to claim 11, further comprising the step of storing in at least one record a relationship that has the URN of a second record as an internal value and storing in said second record a relationship that has the URN of said at least one record as an internal value.

13. The method for representing information in a computer system according to claim 11, further comprising the step of naming each record by storing a designated name of thereof as an external value of a relationship in each record.

14. The method for representing information in a computer system according to claim 11, further comprising the step of providing an index to said knowledge representation database made up of a name of each record together with its associated URN, wherein said name is a designated external value of a relationship stored therein as being the name of a concept represented by that record.

15. The method for representing information in a computer system according to claim 14, further comprising the step of designating certain records as System Concepts in a system concept index to said database, said system concepts being recognized by the system by determining whether the URN or the name of a particular record is in said system concept index.

16. The method for representing information in a computer system according to claim 15, further comprising the step of defining System Concepts designating strata, levels of abstraction, or libraries (such as System, Attribute, Component, and Project).

17. The method for representing information in a computer system according to claims 15, further comprising the step of defining System Concepts designating attribute classes (such as Assignment, Connection, Non-Binding, and External).

18. The method for representing information in a computer system according to claim 17, further comprising the steps of:

reading the relationships stored in a designated active concept record and identifying the attribute class concepts of the characterizations of said relationships;

identifying relevant documents to said record by comparing said identified attribute class concepts to system concepts known by the system to be associated with the types of documents which the system can automatically derive;

selecting from said identified relevant documents a specific document for display of said record;

deriving said specific document; and

displaying said specific document.

19. The method for representing information in a computer system according to claim 18, wherein the step of deriving a specific document includes the steps of:

selecting a view, class, and type of display for said active concept record;

automatically deriving the selected type of display by reading a subset of the records in said knowledge representation database relative to said active concept record, based on stored relationships therein, which subset defines the concepts constituting said selected type of display;

assigning icons to said subset of the records according to said selected type of display;

organizing and locating said icons in a display space of said computer system according to said selected class; and

creating connection icons for interconnections between concept icons located in said display space according to said selected view.

20. The method for representing information in a computer system according to claim 19, wherein the type specifies particular procedures for selecting a subset of the information in said knowledge representation database, reading the descriptive networks for the concepts in said subset, and assigning icons to the concepts in said subset.

21. The method for representing information in a computer system according to claim 19, wherein the class identifies the relevant types and specifies particular procedures for organizing and locating the icons of the concepts in said display space.

22. The method for representing information in a computer system according to claim 19, wherein the view identifies the relevant classes and specifies particular procedures for creating icons for the relationships between the concepts located in the document display space.

23. The method for representing information in a computer system according to claim 15, further comprising the step of defining System Concepts designating attribute properties (such as Name, Data Type, Field Length, and Prompt).

24. The method for representing information in a computer system according to claim 15, further comprising the step of defining system concepts designating pattern storage characterizations (such as Children, Node Name Attr, Parent, Use Isa Name, Concat Attrs, Os Path Name, Sibling Inc, User Text, Connect To, Linear, Multiple, and Single).

25. The method for representing information in a computer system according to claim 15, further comprising the steps of:

evaluating a natural language user input expression by sequential application of an augmented transition network parse based on identification of the function of a recognized expression in said knowledge representation database, and by parsing by differences based on the evaluation of said recognized expression and said identified function;

wherein a recognized expression corresponds to a name associated with a record in the knowledge representation database, and the function of said recognized expression corresponds to the system concepts in a descriptive network of said recognized expressions.

26. The method for representing information in a computer system according to claim 25, wherein the recognition of the user input expression occurs by matching the input expression to names of records in said knowledge representation database and associating with said input expression the library system concept in the parent fundamental relationship of said records.

27. The method for representing information in a computer system according to claim 26, further comprising the steps of dynamically defining concepts not found in said knowledge representation database and adding said defined concepts to the knowledge representation database in a particular location within a particular library through evaluation of said input expressions as a constraint on the fundamental relationships of said dynamically defined concepts.

28. The method for representing information in a computer system according to claim 26, further comprising the steps of coordinating the derivation of appropriate documents from the information in the knowledge representation database in response to said natural language user input expression through evaluation of the input expression for specifying the active concepts and selected appropriate view, class, and type of display for response to the input expression.

29. The method for representing information in a computer system according to claim 11, further comprising the steps of:

establishing, for each record in the knowledge representation database, fundamental relationships between said record and other records of said knowledge representation database, said fundamental relationships being comprised of;

intrastratum relationships which store URNs of other records on the same strata or level of abstraction designated as separate libraries within the knowledge representation system, said intrastratum relationships comprised of parent and children relationships, and

interstrata relationships which store URNs of other records in different strata or libraries, said interstrata relationships being comprised of Type relationships which identify the record in which they are stored as a particular instance of records in another stratum or library.

30. The method for representing information in a computer system according to claim 11, further comprising the step of establishing at least one optional user relationship stored in at least one of the records in the knowledge representation database by specifying a user defined Attribute that is not a system concept and storing the URN of said user defined Attribute in said at least one record as the characterization of said optional user relationship.

31. The method for representing information in a computer system according to claim 11, further comprising the step of establishing in said computer system a plurality of editors for modifying the relationships stored in said records of said knowledge representation database, wherein said editors enable users to identify and store relationships in said records in said knowledge representation database by specifying the characterization and value of a relationship and storing said relationships in at least one of said records.

32. The method for representing information in a computer system according to claim 31, wherein relationships are created further comprising the steps of:

identifying the characterization of a relationship pertinent to a given record;

designating the record in which said characterization is to be stored, wherein said designated record is either said given record or one of its taxonomy ancestors; and

storing said characterization in said designated record in a relationship with a nil value.

33. The method for representing information in a computer system according to claim 31, wherein relationships are edited further comprising the steps of:

identifying the value to be edited;

identifying a source record of said value to be edited, wherein said source record is the record in said knowledge representation database wherein said value to be edited is stored;

modifying said value to be edited to obtain a new value; and

replacing said value to be edited in the relationship in said source record in which said value to be edited is stored with said new value.

34. The method for representing information in a computer system according to claim 31, further comprising the steps of:

establishing in said computer system means for recognizing, storing, and operating on patterns in the records and relationships created by said modifying editors; and

recognizing the patterns in the records and relationships by recording the plurality of URNs of records and relationships created by one of said editors and determining whether said URNs are typical of the operation of said one of said editors in modifying said records and relationships.

35. The method for representing information in a computer system according to claim 34, wherein said patterns in said relationships are comprised of recurring correlations in values of fundamental relationships in Project library records, and recurring correlations in values of user relationships in records of the Project and Component libraries.

36. The method for representing information in a computer system according to claim 35, wherein patterns recognized in Project library records are stored in the component record represented by the URN stored in the type fundamental relationship of said project records or in a parent record of said component record, and wherein patterns recognized in Component library records are stored in the component record in which said recognized patterns are first identified or in a parent record of said component record;

the characterization of the relationship storing the recognized pattern being designated as a system concept.

37. The method for representing information in a computer system according to claim 34, further comprising the steps of:

establishing system concepts to characterize relationships storing values that define recognized patterns established by said one of said editors, wherein each of said system concepts identify particular types of patterns created by said one of said editors; and

storing relationships characterized by said system concepts in relevant records comprising said knowledge representation database.

38. The method for representing information in a computer system according to claim 34, further comprising the steps of:

identifying the relevant relationship storing said stored pattern as a value;

reading said relevant relationship; and

using the pattern stored within said relevant relationship within the operation of said one of said editors.

39. The method for representing information in a computer system according to claim 11, further comprising the step of establishing in said computer system a plurality of editors for that enable users to create a new record in said knowledge representation database, said editors including means for establishing appropriate fundamental relationships of said new record to other records in said knowledge representation database and storing said fundamental relationships in said new record.

40. The method for representing information in a computer system according to claim 11, further comprising the step of:

establishing in said computer system a descriptive database for describing an active concept record designated by a user, comprised of a plurality of records each of which stores a single relationship having an associated URN for said active concept record, and an associated URN for a source record in which said relationship is stored,

the URN for said active concept being the URN of the record in said knowledge representation database for which the description in the descriptive database is assembled, and

the URN for the source record being the URN of that record in the knowledge representation database in which said relationship is stored.

41. The method for representing information in a computer system according to claim 40, further comprising the steps of:

reading a descriptive network for an active concept in the knowledge representation database forming a network of related records through the fundamental relationships of parent and type, combining the relationship lists from said read records, and storing the said relationships from said read records in said descriptive database,

said relationship lists being combined by applying inheritance rules comprised of Taxonomy, Type, Composition and User inheritance,

each relationships comprising said relationship list being stored in said descriptive database together with the URN of said active concept and the URN of source record, wherein said each relationship is stored in the Knowledge Representation Database.

42. The method for representing information in a computer system according to claim 41, wherein Taxonomy inheritance rules define manners of combining the relationship list in the series of records linked through the parent fundamental relationship for component and attribute library records.

43. The method for representing information in a computer system according to claim 41, wherein Type inheritance rules define manners of combining the relationship list in the component records identified by the URNs stored in the type fundamental relationship of a project record.

44. The method for representing information in a computer system according to claim 41, wherein Composition inheritance rules define manners of combining the relationship list in the series of records linked through the parent fundamental relationship for project library records.

45. The method for representing information in a computer system according to claim 41, wherein User inheritance rules define manners of combining the relationship lists in a series of records linked through the URNs stored as internal values in relationships stored in said records.

46. The method for representing information in a computer system according to claim 11, further comprising the steps of:

establishing user interaction with said knowledge representation database through interaction with a display of said computer system;

selecting a view, class, and type of display for a designated active concept record and automatically deriving the selected type of display by reading a subset of the records in said knowledge representation database relative to said active concept, based on stored relationships therein, which subset defines the concepts constituting said selected type of display;

assigning icons to said subset of the records according to said selected type of display;

organizing and locating said icons in the display space of said computer system according to said selected class, and creating connection icons for interconnections between concept icons located in said display space according to said selected view;

detecting user interaction with the icons in said display space;

evaluating said icon interaction through the use of decision trees for evaluation of the view, command history, system flags, and icon association for identification of the appropriate process or procedure for response to said interaction; and

applying said appropriate process or procedure for response and developing a response to said interaction.

47. The method for representing information in a computer system according to claim 46, wherein said icon is interpreted as the entities comprising said icon.

48. The method for representing information in a computer system according to claim 46, wherein said icon is interpreted as the location in said display space occupied by said icon.

49. The method for representing information in a computer system according to claim 46, wherein said icon is interpreted as one of the plurality of URNs associated with said icon.

50. The method for representing information in a computer system according to claim 9, wherein each record is also associated with a name by which each record is identified.

51. The method for representing information in a computer system according to claim 9, further comprising the step of establishing in said computer system a plurality of editors for modifying the records of said knowledge representation database and information stored in said records, wherein said editors enable users to create, read and replace information in each of the records in the knowledge representation database.

52. A method for representing information in a computer system, comprising the steps of:

establishing in said computer system a knowledge representation database made up of individual records, wherein each record is associated with a unique reference number (URN) by which each record is identified and wherein each record stores at least one relationship comprised of a characterization and a value,

the characterization of said relationship being a URN of a second record which defines the nature of said relationship, and

the value of said relationship being a complex data representation composed of at least one internal value, external value, or mixed value which define the object of said relationship,

internal values storing only URNs of other records,

external values storing external data such as character strings, integers, and real numbers, and

mixed values storing a combination of internal and external values;

designating certain records of said knowledge representation database as System Concepts by including the names and associated URNs of said certain records in a System Concept index to said knowledge representation database provided in said computer system;

creating user defined records based on said System Concepts, and storing said user defined records in said knowledge representation database, wherein at least one user defined record stores therein the URN of a System Concept as a relationship thereof, and wherein all of said user defined records store therein the URN of least one other record in said knowledge representation database; and

deriving comprehensible information through evaluation of relationships stored in said records of said knowledge representation database and recognition of System Concepts through consultation of said System Concept index.

53. A method for representing information in a computer system, comprising the steps of:

establishing in said computer system a knowledge representation database made up of individual records, wherein each record is associated with a unique reference number (URN) by which each record is identified and wherein each record stores at least one relationship comprised of a URN to a second record, said stored relationships constituting a knowledge network conceptually interconnecting said records of said database in said computer system so as to yield comprehensible information; and

deriving comprehensible information by designating a record in said database as an active record, assembling in said computer system all records related to said active record by reading relationships stored in said active record, and deriving information about said designated active record by evaluating said assembled records.