Method and apparatus for interpreting data and accessing on-line documentation in a computer system

• US 5,598,511 A
• Filed: 12/28/1992
• Issued: 01/28/1997
• Est. Priority Date: 12/28/1992
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Berg, Erik O., Aguiar, Judy, Petrinjak, Jan S., Corbett, Terri L.
• Primary Examiner(s): Downs, Robert W.
• Application Number:
  US07/999,279
• Time in Patent Office: 1,492 Days
• Field of Search: 395/50, 395/51, 395/54, 395/77, 395/53
• US Class Current: 706/50
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

A modular expert system that uses knowledge modules from multiple personal computer applications to solve a problem. The expert system accesses additional knowledge modules from independent computer applications to reach a conclusion regarding a computer system or application problem. A learning module records past problem-solving histories for reference by the expert system in enhancing problem solving capabilities. A knowledge acquisition module allows an application developer to create knowledge modules for new applications to be used with the modular expert system. A method for solving problems using modular knowledge modules from multiple applications is also described. The method compares a list of facts to rules in its current knowledge base. If a valid conclusion can not be reached then another knowledge module is selected by an inference engine to replace the existing knowledge module to provide further knowledge in solving the problem.

Claims

1. An apparatus for assisting a user in solving computer system and application problems in a computer system which is capable of executing a plurality of independent application programs, the apparatus for assisting a user comprising:

a working memory for storing a list of facts about the problem;

a modular knowledge base coupled to said working memory, said modular knowledge base having rules corresponding to meta knowledge, general knowledge, and a current knowledge module, wherein said meta knowledge and said general knowledge being application-independent, said current knowledge module being specific to a current application program selected for execution by the computer system, and said current knowledge module being replaceable by additional knowledge modules from additional ones of said application programs, wherein each of said additional knowledge modules corresponds to one of said additional application programs, wherein said additional application programs are distinct from said apparatus for assisting a user; and

an inference engine coupled to said working memory and said modular knowledge base, said inference engine for interpreting the facts in said working memory by applying the facts to said rules corresponding to meta knowledge, general knowledge, or said current knowledge module in said modular knowledge base to produce a valid conclusion about solving the problem, wherein when said inference engine is unable to produce a valid conclusion using said current knowledge module said inference engine selects one of said additional knowledge modules corresponding to one of said additional application programs and replaces said current knowledge module with said selected additional knowledge module such that said selected additional knowledge module becomes said current knowledge module.

2. The apparatus of claim 1 wherein when said inference engine cannot produce a valid conclusion using the current knowledge module said inference engine repeatedly selects and replaces said current knowledge module with said additional knowledge modules to further interpret the facts until the valid conclusion is produced.

3. The apparatus of claim 2 wherein when said selected knowledge module becomes said current knowledge module, said inference engine receives additional facts from said application program corresponding to said selected knowledge module.

4. The apparatus of claim 1, further comprising:

a knowledge acquisition module coupled to said modular knowledge base, said knowledge acquisition module for querying the user for the rules, relationships between the rules, and starting information, and checking the logical correctness of the rules and relationships to assist the user in creating a new knowledge module for a new application program.

5. The apparatus of claim 4, further comprising:

a learning module coupled to said modular knowledge base and said inference engine, said learning module for keeping a log of information including the facts and corresponding results and valid conclusions reached by said inference engine, said inference engine for using the log to reach the valid conclusion about the problem.

6. The apparatus of claim 5 further comprising:

a hypertext module coupled to said modular knowledge base and said inference engine, said hypertext module for storing on-line documentation associated with the current application program, the valid conclusion for indexing said hypertext module to identify on-line documentation relevant to the problem.

7. The apparatus of claim 6 wherein said hypertext module stores text, graphics, and multimedia information associated with said current and additional application programs and the valid conclusion indexes said hypertext module to identify text graphics, and multimedia information relevant to the problem.

8. The apparatus of claim 7, further comprising:

a user interface coupled to said inference engine and said hypertext module, said user interface for allowing the user to interact with the operation of said apparatus for assisting a user, said user interface also for allowing the user to determine the list of facts stored in said working memory.

9. The apparatus of claim 8 wherein said inference engine operates in cycles and analyzes information by forward reasoning.

10. The apparatus of claim 1 wherein said working memory receives the facts from said current application program.

11. In a computer system having the ability to execute a plurality of independent application programs, a method for interpreting data from one or more of the computer application programs and from previous test results stored in a learning module to assist a user in solving a computer system and application problem, the method comprising the computer implemented steps of:

storing meta knowledge and general knowledge in a knowledge base, said meta knowledge being rules of logic, said general knowledge being application-independent rules;

loading a first knowledge module into the knowledge base, said first knowledge module being specific to a first application program;

receiving a list of facts about the problem;

comparing a list of facts to the meta knowledge, general knowledge, or first knowledge module rules in the knowledge base, wherein said comparing the list of facts produces additional facts by application of the facts to the meta knowledge, general knowledge, or first knowledge module rules in the knowledge base, said comparing the list of facts being repeated until either a valid conclusion is reached or no additional facts are produced;

determining whether the learning module stores previous test results that are relevant to solving the problem;

comparing the previous test results that are relevant to solving the problem to the meta knowledge, general knowledge, or first knowledge module rules in the knowledge base, wherein said comparing the previous test results produces additional facts by application of the previous test results to the meta knowledge, general knowledge, or first knowledge module rules in the knowledge base, said comparing the previous test results being repeated until either a valid conclusion is reached or no additional facts are produced; and

selecting a second knowledge module which provides additional information about solving the problem, wherein said second knowledge module being specific to a second application program, the first and second application programs being independent;

wherein said comparing the list of facts, comparing the previous test results, and selecting a second knowledge module being performed by an inference engine, wherein said inference engine interprets the list of facts by applying the list of facts to the meta knowledge, general knowledge, or second knowledge module rules in the knowledge base to solve the problem.

12. The method of claim 11, further comprising the step of:

replacing said first knowledge module with the second knowledge module when said comparing the list of facts does not result in a valid conclusion; and

repeating said comparing the list of facts and comparing the previous test results until either a valid conclusion is reached or no additional facts are produced.

13. The method of claim 12, further comprising the steps of:

accessing on-line documentation; and

indexing said on-line documentation corresponding to the valid conclusion.