Expert system knowledge-deficiency reduction through automated database updates from semi-structured natural language documents

• US 6,374,261 B1
• Filed: 04/09/1999
• Issued: 04/16/2002
• Est. Priority Date: 04/09/1999
• Status: Expired due to Term

• Current Assignee: Avaya Incorporated
• Original Assignee: Avaya Technology Corporation Miami Lakes FLA US
• Inventors: Prabhakar, Gokul Chander, Alvarez, Miguel O.
• Primary Examiner(s): Alam, Hosain T.
• Assistant Examiner(s): NGUYEN, TAM V
• Application Number:
  US09/289,682
• Time in Patent Office: 1,103 Days
• Field of Search: 706/45, 706/50, 706/60, 707/200
• US Class Current: 706/45
• CPC Class Codes: G06N 5/025 Extracting rules from data

Abstract

Knowledge deficiency of an expert system (110) is alleviated by automating extraction of data that is relevant to the expert system'"'"'s diagnostics knowledge database (112) and automatically and in an on-going manner updating the knowledge database with the extracted information. Evolving diagnostic knowledge in a domain is represented in the form of semi-structured natural-language reports (116). A computer-implemented tool (120) analyzes (206) a report to determine its kind, and then employs (222) an intelligent filter (124,126) corresponding to that kind of report to parse (224-244) the report'"'"'s sections, to extract (226-244) data therefrom, and to perform consistency checks (228, 234) between sections on the extracted data. The computer then assembles (248) the extracted data into database records and includes (250) them in the knowledge database.

Claims

1. A method of automatically updating a knowledge base comprising a database that stores information pertaining to a subject, comprising:

in response to obtaining a file of information that is at least in part expressed in natural-language form, analyzing the file by computer to identify therein types of information pertaining to the subject;

analyzing by computer the information of the identified types by executing an intelligent filter that uses heuristics to identify therein items of information for storage in the knowledge database;

extracting by computer the found items of information from the file;

arranging by computer the extracted items by their types into a database record; and

storing by computer the database record in the knowledge database.

2. The method claim 1 wherein:

the knowledge database is a knowledge database of an expert system;

the subject comprises a target of the expert system; and

the file contains information pertaining to the subject.

3. The method of claim 2 wherein:

the subject comprises a plurality of targets of the expert system;

the step of arranging comprises

determining from an extracted item which ones of the plurality of the targets the file pertains to, and

creating a separate database record for each of the ones of the plurality of the targets; and

the step of storing comprises

storing each of the created database records in the knowledge database.

4. The method of claim 1 wherein:

the file is a semi-structured data entity, comprising a plurality of predefined types of information at least some of which comprise information expressed in natural language form.

5. The method of claim 1 wherein:

each analyzing comprises parsing by computer.

6. The method of claim 1 further comprising:

checking an identified item of information for consistency between a plurality of the identified types of information; and

in response to determining that the identified item of information is not consistent between the plurality of identified types of information, generating a warning.

7. The method of claim 1 wherein:

the file of information is either a first or a second kind of file; and

analyzing the file comprises

analyzing the file by computer to determine the kind of the file,

in response to determining that the file is of the first kind, analyzing the file and the information by a first program-implemented information filter, and

in response to determining that the file is of the second kind, analyzing the file and the information by a second program-implemented information filter.

8. The method of claim 1 wherein:

analyzing by computer the information of the identified types comprises

parsing the information of a said identified type using keywords or phrases.

9. The method of claim 1 wherein:

analyzing by computer the information of the identified types comprises

parsing the information of a said identified type by executing said intelligent filter that uses keyword-based or phrase-based heuristics.

10. An apparatus that performs the method of claim 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9.

11. A computer-readable medium containing instructions which, when executed in a computer, cause the computer to perform the method of claim 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9.