Entity structured representation and variant generation

1. A computer-implemented method comprising:

automatically generating an active learning-based parser by (i) learning one or more semantic components attributed to a first set of labeled instances of a given entity type and (ii) mapping a second set of unlabeled instances of the given entity type to at least one of the one or more semantic components;

automatically parsing, from input text, at least a portion of a third set of unlabeled instances of the given entity type into at least one of the one or more semantic components, wherein said automatically parsing comprises implementing the active learning-based parser based at least in part on one or more items of information relevant to the given entity type;

generating, based at least in part on (i) the one or more semantic components and (ii) information pertaining to one or more valid component-specific variants, one or more variants of the one or more semantic components;

creating, based at least in part on the one or more variants of the one or more semantic components, one or more variants of at least one instance of an entity associated with the given entity type in the input text; and

outputting, to at least one user, the one or more variants of the at least one instance of the entity;

wherein the method is carried out by at least one computing device.

2. The computer-implemented method of claim 1 , wherein the information pertaining to one or more valid component-specific variants comprises user-provided information.

3. The computer-implemented method of claim 1 , comprising:

generating the active learning-based parser such that the active learning-based parser maps (i) the information pertaining to one or more valid component-specific variants of a particular instance of an entity to (ii) one or more structural aspects of the particular instance of an entity.

4. The computer-implemented method of claim 1 , comprising:

predicting one or more structural aspects of one or more unlabeled instances of the entity.

5. The computer-implemented method of claim 4 , comprising:

outputting, to a user for review, the predicted structural aspects of the one or more unlabeled instances.

6. The computer-implemented method of claim 5 , comprising:

integrating feedback from the user, in response to the review of the predicted structural aspects of the one or more unlabeled instances, into the model.

7. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:

automatically generate an active learning-based parser by (i) learning one or more semantic components attributed to a first set of labeled instances of a given entity type and (ii) mapping a second set of unlabeled instances of the given entity type to at least one of the one or more semantic components;

automatically parse, from input text, at least a portion of a third set of unlabeled instances of the given entity type into at least one of the one or more semantic components, wherein said automatically parsing comprises implementing the active learning-based parser based at least in part on one or more items of information relevant to the given entity type;

generate, based at least in part on (i) the one or more semantic components and (ii) information pertaining to one or more valid component-specific variants, one or more variants of the one or more semantic components;

create, based at least in part on the one or more variants of the one or more semantic components, one or more variants of at least one instance of an entity associated with the given entity type in the input text; and

output, to at least one user, the one or more variants of the at least one instance of the entity.

8. The computer program product of claim 7 , wherein the information pertaining to one or more valid component-specific variants comprises user-provided information.

9. The computer program product of claim 7 , wherein the program instructions executable by a computing device further cause the computing device to:

generate the active learning-based parser such that the active learning-based parser maps (i) the information pertaining to one or more valid component-specific variants of a particular instance of an entity to (ii) one or more structural aspects of the particular instance of an entity.

10. The computer program product of claim 7 , wherein the program instructions executable by a computing device further cause the computing device to:

predict one or more structural aspects of one or more unlabeled instances of the entity.

11. The computer program product of claim 10 , wherein the program instructions executable by a computing device further cause the computing device to:

output, to a user for review, the predicted structural aspects of the one or more unlabeled instances.

12. The computer program product of claim 11 , wherein the program instructions executable by a computing device further cause the computing device to:

integrate feedback from the user, in response to the review of the predicted structural aspects of the one or more unlabeled instances, into the model.

13. A system comprising:

a memory; and

at least one processor operably coupled to the memory and configured for:

automatically generating an active learning-based parser by (i) learning one or more semantic components attributed to a first set of labeled instances of a given entity type and (ii) mapping a second set of unlabeled instances of the given entity type to at least one of the one or more semantic components;

automatically parsing, from input text, at least a portion of a third set of unlabeled instances of the given entity type into at least one of the one or more semantic components, wherein said automatically parsing comprises implementing the active learning-based parser based at least in part on one or more items of information relevant to the given entity type;

generating, based at least in part on (i) the one or more semantic components and (ii) information pertaining to one or more valid component-specific variants, one or more variants of the one or more semantic components;

creating, based at least in part on the one or more variants of the one or more semantic components, one or more variants of at least one instance of an entity associated with the given entity type in the input text; and

outputting, to at least one user, the one or more variants of the at least one instance of the entity.

14. A computer-implemented method comprising:

automatically generating an active learning-based parser by (i) learning one or more semantic components attributed to a first set of labeled instances of a given entity type and (ii) mapping a second set of unlabeled instances of the given entity type to at least one of the one or more semantic components;

automatically parsing, from input text, at least a portion of a third set of unlabeled instances of the given entity type into at least one of the one or more semantic components by implementing the active learning-based parser based at least in part on one or more items of information relevant to the given entity type;

generating, based at least in part on (i) the one or more semantic components and (ii) information pertaining to one or more valid component-specific variants, multiple variants of the one or more semantic components;

outputting, to at least one user for user review, the multiple variants of the one or more semantic components;

creating, based at least in part on feedback from the user review of the multiple variants of the one or more semantic components, one or more variants of at least one instance of an entity associated with the given entity type in the input text; and

outputting, to the at least one user, the one or more variants of the at least one instance of the entity;

wherein the method is carried out by at least one computing device.

15. The computer-implemented method of claim 14 , wherein the information pertaining to one or more valid component-specific variants comprises user-provided information.

16. The computer-implemented method of claim 14 , comprising:

generating the active learning-based parser such that the active learning-based parser maps (i) the information pertaining to one or more valid component-specific variants of a particular instance of an entity to (ii) one or more structural aspects of the particular instance of an entity.