Generating rules for automated text annotation

Patent metadata

Field	Value
Publication number	US-10902198-B2
Application number	US-201816203903-A
Country	US
Kind code	B2
Filing date	Nov 29, 2018
Priority date	Nov 29, 2018
Publication date	Jan 26, 2021
Grant date	Jan 26, 2021

Abstract

Natural language text and annotated text can be received. The annotated text can specify at least one anchor and at least one trigger contained in the natural language text and indicate a correspondence between the anchor and the trigger. The natural language text, the annotated text and at least one parse tree generated from the natural language text can be processed. Based on the processing, at least one natural language processing rule can be generated and output. The natural language processing rule can be configured to be executed by a processor to process other natural language text.

First claim

What is claimed is:

1. A method comprising:
receiving a first natural language text and a first annotated text, the first annotated text specifying at least one anchor and at least one trigger contained in the first natural language text and indicating a correspondence between the anchor and the trigger;
processing, using a processor, the first natural language text, the first annotated text and at least one parse tree generated from the first natural language text and, based on the processing, generating at least one natural language processing rule, the natural language processing rule configured to be executed by at least a second processor to process at least a second natural language text; and
outputting the at least one natural language processing rule.

2. The method of claim 1 , wherein the natural language processing rule is configured to be executed by the second processor to identify at least one other anchor and at least one other trigger contained in the second natural language text.

3. The method of claim 2 , wherein the natural language processing rule further is configured to be executed by the second processor to identify a correspondence between the at least one other anchor and the at least one other trigger contained in the second natural language text.

4. The method of claim 1 , wherein the correspondence is a semantic relationship between the at least one anchor and the at least one trigger contained in the first natural language text.

5. The method of claim 1 , further comprising:
generating at least one heuristic based on the processing the first natural language text, the first annotated text and the at least one parse tree;
wherein generating the at least one natural language processing rule comprises generating the at least one natural language processing rule in accordance with the at least one heuristic.

6. The method of claim 1 , further comprising:
validating the natural language processing rule by applying the natural language processing rule to a plurality of natural language test samples for which anchor and trigger semantic relationships are known; and
generating validation results based on the validating.

7. The method of claim 1 , further comprising:
generating the parse tree from the first natural language text.

8. A system, comprising:
a processor programmed to initiate executable operations comprising:
receiving a first natural language text and a first annotated text, the first annotated text specifying at least one anchor and at least one trigger contained in the first natural language text and indicating a correspondence between the anchor and the trigger;
processing the first natural language text, the first annotated text and at least one parse tree generated from the first natural language text and, based on the processing, generating at least one natural language processing rule, the natural language processing rule configured to be executed by at least a second processor to process at least a second natural language text; and
outputting the at least one natural language processing rule.

9. The system of claim 8 , wherein the natural language processing rule is configured to be executed by the second processor to identify at least one other anchor and at least one other trigger contained in the second natural language text.

10. The system of claim 9 , wherein the natural language processing rule further is configured to be executed by the second processor to identify a correspondence between the at least one other anchor and the at least one other trigger contained in the second natural language text.

11. The system of claim 8 , wherein the correspondence is a semantic relationship between the at least one anchor and the at least one trigger contained in the first natural language text.

12. The system of claim 8 , the executable operations further comprising:
generating at least one heuristic based on the processing the first natural language text, the first annotated text and the at least one parse tree;
wherein generating the at least one natural language processing rule comprises generating the at least one natural language processing rule in accordance with the at least one heuristic.

13. The system of claim 8 , the executable operations further comprising:
validating the natural language processing rule by applying the natural language processing rule to a plurality of natural language test samples for which anchor and trigger semantic relationships are known; and
generating validation results based on the validating.

14. The system of claim 8 , the executable operations further comprising:
generating the parse tree from the first natural language text.

15. A computer program product, comprising:
a computer readable storage medium having program code stored thereon, the program code executable by a data processing system comprising a first processor to initiate operations including:
receiving a first natural language text and a first annotated text, the first annotated text specifying at least one anchor and at least one trigger contained in the first natural language text and indicating a correspondence between the anchor and the trigger;
processing the first natural language text, the first annotated text and at least one parse tree generated from the first natural language text and, based on the processing, generating at least one natural language processing rule, the natural language processing rule configured to be executed by at least a second processor to process at least a second natural language text; and
outputting the at least one natural language processing rule.

16. The computer program product of claim 15 , wherein the natural language processing rule is configured to be executed by the second processor to identify at least one other anchor and at least one other trigger contained in the second natural language text.

17. The computer program product of claim 16 , wherein the natural language processing rule further is configured to be executed by the second processor to identify a correspondence between the at least one other anchor and the at least one other trigger contained in the second natural language text.

18. The computer program product of claim 15 , wherein the correspondence is a semantic relationship between the at least one anchor and the at least one trigger contained in the first natural language text.

19. The computer program product of claim 15 , wherein the program code is executable by the data processing system to initiate operations further comprising:
generating at least one heuristic based on the processing the first natural language text, the first annotated text and the at least one parse tree;
wherein generating the at least one natural language processing rule comprises generating the at least one natural language processing rule in accordance with the at least one heuristic.

20. The computer program product of claim 15 , wherein the program code is executable by the data processing system to initiate operations further comprising:
validating the natural language processing rule by applying the natural language processing rule to a plurality of natural language test samples for which anchor and trigger semantic relationships are known; and
generating validation results based on the validating.

Assignees

• IBM

Inventors

• Doganata Yurdaer N
• Delima Roberto
• Ezen Can Aysu

Classifications

• G06F40/30 Primary

  Semantic analysis · CPC title

• G06F40/205 Primary

  Parsing · CPC title

• G10L15/1815

  Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning · CPC title

• G06F40/169

  Annotation, e.g. comment data or footnotes · CPC title