Method and device for processing documents and recording medium

• US 7,047,486 B1
• Filed: 01/18/2000
• Issued: 05/16/2006
• Est. Priority Date: 01/21/1999
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Nagao, Katashi
• Primary Examiner(s): Shah, Sanjiv
• Application Number:
  US09/623,632
• Time in Patent Office: 2,310 Days
• Field of Search: 715/514, 715/500, 707/3, 707/5, 707/6, 707/102, 707/104.1
• US Class Current: 715/222
• CPC Class Codes: G06F 16/353 into predefined classes
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

A method and an apparatus for processing a document of a tagged internal structure made up of a plurality of elements. A plurality of documents received by a receiving unit are stored in a RAM of a storage unit provided in a main body portion of the apparatus. The characteristic information representing the characteristics of a document is extracted in accordance with the sequence of operations recorded on a ROM under control by controller. Each document is classified into classification items making up a classification model, depending on the degree of interrelation between the characteristic information of the document extracted by a characteristic information extraction unit and the classification-item-based information to realize automatic document classification in such a manner as to reflect the interest of a user.

Claims

1. A method for processing a document having an internal structure made up of a plurality of elements, the method comprising:

a characteristic information extracting step of extracting characteristics of the document;

a document classification step of classifying the document into a plurality of classification items making up a classification model depending on a degree of interrelation between the characteristic information of the document extracted by the characteristic information extracting step and characteristic information of the plurality of classification items; and

a calculation step of calculating the degree of interrelation between the characteristic information of the document extracted in the characteristic information extracting step and the characteristic information of the classification items and classifying the document in one of the classification items with a maximum degree of interrelation when the degree of interrelation exceeds a threshold value.

2. The document processing method according to claim 1, further comprising a step of receiving a plurality of documents, the characteristic information extracting step extracting the characteristic information representing characteristics of each of the plurality of documents received in the step of receiving.

3. The document processing method according to claim 1, wherein a weight is imparted to each of the plurality of elements based on the internal structure of the document, the characteristic information extracting step extracting any element of the plurality of elements having a respective weight larger than a pre-set value.

4. The document processing method according to claim 1 wherein the document classification step classifies the document using a classification model which is prepared based on the classification step of classifying the document into the plurality of classification items.

5. The document processing method according to claim 1, wherein the degree of interrelation of the characteristic information of the document extracted in the characteristic information extracting step and the characteristic information of said the plurality of classification items is a linear linkage of a number of overlapping ones of proper nouns contained in the document characteristic information and of proper nouns contained in the characteristic information of said the plurality of classification items, and a degree of interrelation between semantics other than the proper nouns contained in the document characteristic information and semantics other than the proper nouns contained in said the plurality of classification items.

6. The document processing method according to claim 5, wherein, to obtain the degree of interrelation of the semantics, each of the semantics is weighted based on a structure of reciprocal referencing among the semantics, and a difference of a weight, imparted to one of the semantics as a dependent variable, is divided by a difference of a weight, imparted to the other semantics as an independent variable, with a resulting quotient being the degree of interrelation between the one and the other semantics.

7. The document processing method according to claim 6, wherein the interrelation of reciprocal referencing among the semantics is prepared using a dictionary which, for each of the semantics, references other semantics.

8. A method for processing a document having an internal structure made up of a plurality of semantics, the method comprising:

a weighting step for weighting each of the semantics based on a structure of reciprocal referencing among the semantics, wherein the weighting step imparts a weight based on the structure of reciprocal referencing obtained on organizing the reciprocal referencing of semantics using a dictionary which, for each of the semantics, references other semantics; and

a degree of interrelation calculating step of dividing a difference of a weight, imparted to one of the semantics as a dependent variable, by a difference of a weight, imparted to the other semantics as an independent variable, with a resulting quotient being the degree of interrelation between the one and the other semantics.

9. The document processing method according to claim 8, wherein attribute information representing attributes of each of the semantics of said the dictionary is imparted to each of the semantics in the dictionary, and wherein the structure of the reciprocal referencing is based on the attribute information.

10. A method for processing a document having an internal structure made up of a plurality of elements, the method comprising:

a characteristic information extracting step of extracting characteristics of the document; and

a document classification step of classifying the document into a plurality of classification items making up a classification model depending on a degree of interrelation between the characteristic information of the document extracted by the characteristic information extracting step and the characteristic information of the plurality of classification items,

wherein the degree of interrelation of the characteristic information of the document extracted in the characteristic information extracting step and the characteristic information of the plurality of classification items is a linear linkage of a number of overlapping ones of proper nouns contained in the document characteristic information and of proper nouns contained in the characteristic information of the classification items, and a degree of interrelation between semantics other than the proper nouns contained in the document characteristic information and semantics other than the proper nouns contained in the plurality of classification items.

11. The document processing method according to claim 10, further comprising a step of receiving a plurality of documents, the characteristic information extracting step extracting the characteristic information representing characteristics of each of the plurality of documents received in the step of receiving.

12. The document processing method according to claim 10, wherein a weight is imparted to each of the plurality of elements based on the internal structure of the document, the characteristic information extracting step extracting any element of the plurality of elements having a respective weight larger than a pre-set value.

13. The document processing method according to claim 10 wherein the document classification step classifies the document using a classification model which is prepared based on the classification step of classifying the document into the plurality of classification items.

14. The document processing method according to claim 10 wherein to obtain the degree of interrelation of the semantics, each of the semantics is weighted based on a structure of reciprocal referencing among the semantics, and a difference of a weight, imparted to one of the semantics as a dependent variable, is divided by a difference of a weight, imparted to the other semantics as an independent variable, with a resulting quotient being the degree of interrelation between the one and the other semantics.

15. The document processing method according to claim 14, wherein the interrelation of reciprocal referencing among the semantics is prepared using a dictionary which, for each of the semantics, references other semantics.