DOCUMENT PROCESSING, TEMPLATE GENERATION AND CONCEPT LIBRARY GENERATION METHOD AND APPARATUS

• US 20110295595A1
• Filed: 05/31/2011
• Published: 12/01/2011
• Est. Priority Date: 05/31/2010
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Li, Zhong, Ren, Yong Liang, Wang, Xiao Yuan, Li, Jing, Hu, Gang, Cao, Feng
• Granted Patent: US 8,949,108 B2
• Time in Patent Office: Days
• US Class Current: 704/9
• CPC Class Codes: G06F 40/143 Markup, e.g. Standard Generalized Markup Language [SGML] or Document Type Definition [DTD]
• CPC Class Codes: G06F 40/186 Templates
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

The present invention relates to document processing method and apparatus which can edit a natural language and generate a machine-processable document; a template generating method and apparatus which can be used for document processing method and apparatus; a concept library generating method and apparatus which can be used for the document processing method and apparatus and the template generating method and apparatus. The present invention provided a possibility for semantic interaction of documents in different systems and enhances efficiency.

Claims

1. A document processing method, comprising the steps of:

acquiring a template consisting of at least one semantic item, wherein (i) candidate values of said semantic item and relations among said semantic items are associated with machine-processable semantic codes, and (ii) said template is associated with a corresponding natural language expression mode;

receiving an input value for said semantic item

generating a machine-processable document comprising (i) said machine-processable semantic codes based on said input value, (ii) the relation among said semantic items and (iii) the machine-processable semantic codes as associated; and

generating a natural language document comprising said input value based on said input value and said natural language expression mode.

2. The method of claim 1, wherein receiving said input value comprises:

displaying an input interface of said semantic item; and

receiving said input value via said input interface.

3. The method of claim 2, wherein displaying said input interface comprises:

identifying a core concept in said template;

identifying an interrelation between said semantic item and said core concept in said template;

acquiring a qualifier concept from a concept library, wherein an interrelation between said qualifier concept and said core concept is identical to the interrelation between said semantic item and said core concept; and

generating said input interface for inputting said input value of said semantic item based on said qualifier concept.

4. The method of claim 1, wherein receiving said input value comprises receiving said input value from an external apparatus.

5. The method of claim 1, wherein acquiring said template comprises acquiring said template from an external apparatus.

6. The method of claim 1, wherein acquiring said template comprises constructing said template according to a natural language pattern as required.

7. The method of claim 1, wherein acquiring said template comprises replacing a variable sentence element in a natural language document sample with a semantic item of said template to form said template.

8. The method of claim 1, wherein said document comprises a document based on a clinical document architecture (CDA) standard.

9. The method of claim 1, wherein said template is generated from a natural language document sample by a method comprising the steps of:

identifying a core concept and sentence elements in said natural language document sample;

determining a relation between a current concept of said sentence elements and said core concept;

searching in said concept library for said qualifier concept, wherein a relation between said qualifier concept and said core concept is identical to the relation between said current concept and said core concept;

generating said semantic item of said template based on said searching result; and

generating said candidate value of said semantic item based on the qualifier concept having the identical relation, wherein said qualifier concept having the identical relation exists.

10. The method of claim 9, wherein generating said semantic item comprises:

selecting an interested sentence element as the semantic item having candidate value options of said template from sentence elements for which exist the qualifier concept having the identical relation; and

replacing other sentence elements with semantic items having no candidate value option.

11. The method of claim 1, wherein said template is generated from a natural language document sample, by a method comprising the steps of:

displaying said natural language document sample;

selecting interested sentence elements in said natural language document sample as said core concept and semantic items of said template;

determining a relation between a current concept of said semantic item and said core concept;

searching in said concept library for said qualifier concept, a relation between said qualifier concept and said core concept being identical to the relation between said current concept and said core concept; and

generating said candidate value of said semantic item based on said qualifier concept having the identical relation in the case that said qualifier concept having the identical relation exists.

12. The method of claim 9, wherein generating said candidate value comprises taking an interested concept among the qualifier concept and the current concept as said candidate value of said semantic item.

13. The method of claim 3, wherein generating a concept library comprises:

determining said core concept and said qualifier concept based on said natural language document sample;

determining a relation between said qualifier concept and said core concept; and

storing in association;

(i) said core concept, (ii) said qualifier concept, (iii) the relation between said qualifier concept and said core concept, (iv) corresponding semantic codes.

14. A document processing device, comprising:

means configured to acquire a template consisting of at least one semantic item, wherein (i) candidate values of said semantic items and relations among said semantic items are associated with machine-processable semantic codes, and (ii) said template is associated with a corresponding natural language expression mode;

semantic item input means configured to receive an input value for said semantic item;

machine-processable document generating means configured to generate a machine-processable document comprising (i) said machine-processable semantic codes based on said input value, (ii) the relation among said semantic items, and (iii) said machine-processable semantic codes as associated; and

natural language document generating means configured to generate a natural language document comprising said input value based on said input value and said natural language expression mode.

15. The device of claim 14, wherein said semantic item input means comprises:

means configured to display an input interface of said semantic item; and

means configured to receive said input value via said input interface.

16. The device of claim 15, wherein means for displaying said input interface comprises:

means configured to identify a core concept in said template;

means configured to identify an interrelation between said semantic item and said core concept in said template;

means configured to acquire a qualifier concept from a concept library, wherein an interrelation between said qualifier concept and said core concept is identical to the interrelation between said semantic item and said core concept; and

means configured to generate said input interface for inputting said input value of said semantic item based on said qualifier concept.

17. The device of claim 14, wherein means for generating a template from a natural language document sample comprises:

means configured to identify a core concept and sentence elements in said natural language document sample;

means configured to determine a relation between a current concept of said sentence elements and said core concept;

means configured to search in said concept library for said qualifier concept, wherein a relation between said qualifier concept and said core concept is identical to the relation between said current concept and said core concept;

means configured to generate said semantic item of said template based on said searching result; and

means configured to generate said candidate value of said semantic item based on said qualifier concept having the identical relation, wherein said qualifier concept having the identical relation exists.

18. The device of claim 14, wherein means for generating a template from a natural language document sample comprises:

means configured to display said natural language document sample;

means configured to select interested sentence elements in said natural language document sample as said core concept and semantic items of said template;

means configured to determine a relation between a current concept of said semantic item and said core concept;

means configured to search in said concept library for said qualifier concept, a relation between said qualifier concept and said core concept being identical to the relation between said current concept and said core concept; and

means configured to generate said candidate value of said semantic item based on said qualifier concept having the identical relation wherein said qualifier concept having the identical relation exists.

19. The device of claim 16, wherein generating a concept library comprises:

means configured to determine said core concept and said qualifier concept based on said natural language document sample;

means configured to determine a relation between said qualifier concept and said core concept; and

means configured to store in association;

(i) said core concept, (ii) said qualifier concept, (iii) said relation between said qualifier concept and said core concept, and (iv) corresponding semantic codes.

20. A non-transitory computer readable article of manufacture tangibly embodying computer readable instructions which when executed causes a computer to carry out a document processing method, comprising the steps of:

acquiring a template consisting of at least one semantic item, wherein (i) candidate values of said semantic item and relations among said semantic items are associated with machine-processable semantic codes, and (ii) said template is associated with a corresponding natural language expression mode;

receiving an input value for said semantic item

generating a machine-processable document comprising (i) said machine-processable semantic codes based on said input value, (ii) the relation among said semantic items and (iii) the machine-processable semantic codes as associated; and

generating a natural language document comprising said input value based on said input value and said natural language expression mode.