Apparatus and method accumulating cases to be learned

• US 6,549,752 B2
• Filed: 09/24/2001
• Issued: 04/15/2003
• Est. Priority Date: 01/29/2001
• Status: Expired due to Fees

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Tsukamoto, Koji
• Primary Examiner(s): Rovnak, John Edmund
• Application Number:
  US09/960,298
• Publication Number:
  US 20020102526A1
• Time in Patent Office: 568 Days
• Field of Search: 434/155, 434/156, 434/157, 434/335, 434/362, 706/8, 706/52
• US Class Current: 434/362
• CPC Class Codes: G09B 7/00 Electrically-operated teaching apparatus or devices working with questions and answers mechanically operated G09B3/00; computing arrangements G06F

Abstract

An inference rule for determining a category is generated from a set of correct answer cases registered to a database, and the category of an unknown case is inferred. When a user determines whether or not an inference result is correct and inputs a determination result, the determined unknown case is registered to the database as a new correct answer case. By repeating such a process, correct answer cases are accumulated.

Claims

1. A case accumulating apparatus, comprising:

a storage device storing information about a set of correct answer cases;

a learning device generating an inference rule while referencing the information stored in said storage device, and inferring a target characteristic from a known characteristic of a case to be inferred in compliance with the inference rule;

an inquiry device inquiring of a user as to whether or not an inference result of said learning device is correct, and receiving a response from the user; and

a control device determining the target characteristic of the case to be inferred based on the response, and adding information about the case to be inferred including the determined target characteristic to the set of correct answer cases.

2. The case accumulating apparatus according to claim 1, wherein

said learning device generates a new inference rule based on the information about the set of correct answer cases, to which the information about the case to be inferred is added, and infers a target characteristic of another case in compliance with the new inference rule.

3. The case accumulating apparatus according to claim 1, further comprising

a selecting device selecting an inference result of a case to be inquired of the user from among inference results of a plurality of cases to be inferred.

4. The case accumulating apparatus according to claim 1, wherein

said inquiry device inquires of a plurality of users via a communications network as to whether or not inference results of a plurality of cases to be inferred are correct.

5. A case accumulating apparatus, comprising:

a storage device storing, for each of a plurality of cases, a known characteristic, a target characteristic, and flag information indicating whether or not an inquiry to a user is necessary;

a learning device generating an inference rule from a set of cases having flag information indicating that an inquiry is unnecessary while referencing the information stored in said storage device, obtaining a case to be inferred which has flag information indicating that an inquiry is necessary while referencing the information stored in said storage device, and inferring a target characteristic from a known characteristic of the case to be inferred in compliance with the inference rule;

an inquiry device inquiring of the user as to whether or not an inference result of said learning device is correct, and receiving a response from the user; and

a control device determining the target characteristic of the case to be inferred based on the response, and changing the flag information of the case to be inferred to flag information indicating that an inquiry is unnecessary.

6. A case accumulating apparatus, comprising:

a storage device storing information about a set of documents having categories;

a learning device generating an inference rule while referencing the information stored in said storage device, and inferring a category from a characteristic of a document to be inferred in compliance with the inference rule;

an inquiry device inquiring of a user as to whether or not an inference result of said learning device is correct, and receiving a response from the user; and

a control device determining a category of the document to be inferred based on the response, and adding information about the document to be inferred including the determined category to the information about the set of documents.

7. A case accumulating apparatus, comprising:

a storage device storing information about a set of tagged text data;

a learning device generating an inference rule while referencing the information stored in said storage device, and inferring a category of a tag from a characteristic of text data to be inferred in compliance with the inference rule;

an inquiry device inquiring of a user as to whether or not an inference result of said learning device is correct, and receiving a response from the user; and

a control device determining the category of the tag of the text data to be inferred based on the response, and adding information about the text data to be inferred including the determined category to the set of text data.

8. A case accumulating apparatus, comprising:

a storage device storing information about a set of images;

a learning device generating an inference rule while referencing the information stored in said storage device, and inferring a target characteristic from a known characteristic of an image to be inferred in compliance with the inference rule;

an inquiry device inquiring of a user as to whether or not an inference result of said learning device is correct, and receiving a response from the user; and

a control device determining the target characteristic of the image to be inferred based on the response, and adding information about the image to be inferred including the determined target characteristic to the set of images.

9. A computer-readable storage medium on which is recorded a program for causing a computer to execute a process, the process comprising:

generating an inference rule while referencing information about a set of correct answer cases;

inferring a target characteristic from a known characteristic of a case to be inferred in compliance with the inference rule;

inquiring of a user as to whether or not an inference result is correct;

receiving a response from the user;

determining the target characteristic of the case to be inferred based on the response; and

adding information about the case to be inferred including the determined target characteristic to the set of correct answer cases.

10. A propagation signal for propagating a program to a computer, the program causing the computer to execute:

generating an inference rule while referencing information about a set of correct answer cases;

inferring a target characteristic from a known characteristic of a case to be inferred in compliance with the inference rule;

inquiring of a user as to whether or not an inference result is correct;

receiving a response from the user;

determining the target characteristic of the case to be inferred based on the response; and

adding information about the case to be inferred including the determined target characteristic to the set of correct answer cases.

11. A case accumulating method, comprising:

registering information about a set of correct answer cases to a computer;

generating an inference rule while referencing registered information;

inferring a target characteristic from a known characteristic of a case to be inferred in compliance with the inference rule;

inquiring of a user as to whether or not an inference result is correct;

receiving a response from the user;

determining the target characteristic of the case to be inferred based on the response; and

adding information about the case to be inferred including the determined target characteristic to the set of correct answer cases.

12. A case accumulating apparatus, comprising:

storage means for storing information about a set of correct answer cases;

learning means for generating an inference rule while referencing the information stored in said storage means, and for inferring a target characteristic from a known characteristic of a case to be inferred;

inquiry means for inquiring of a user as to whether or not an inference result of said learning means is correct, and for receiving a response from the user; and

control means for determining the target characteristic of the case to be inferred based on the response, and for adding information about the case to be inferred including the determined target characteristic to the set of correct answer cases.