Automated reading comprehension

1. A computer-implemented method, comprising:

identifying, by a computing system based at least in part on a segment of text, a first entity and a second entity that is distinct from the first entity;

determining, by the computing system based at least in part on the segment of text, that a first entity attribute associated with the first entity in the segment of text and a second entity attribute associated with the second entity in the segment of text share an attribute class;

in response to the determining, providing, by the computing system, an indication that, in the segment of text, the first entity attribute is either similar to or different than the second entity attribute; and

generating, by the computing system based on an answer choice that directly describes a relationship between the first entity and the first entity attribute and indirectly describes a relationship between the second entity and the second entity attribute, a textual rewrite that directly describes a relationship between the second entity and the second entity attribute.

2. The computer-implemented method of claim 1 , further comprising comparing the textual rewrite to textual rewrites generated based on the segment of text to determine a veracity of the answer choice.

3. The computer-implemented method of claim 1 , wherein the identifying the first and second entities is based on a question posed about the segment of text.

4. The computer-implemented method of claim 1 , further comprising opting to provide either the indication that the first entity attribute is similar to the second entity attribute or different than the second entity attribute based on a type of question posed about the segment of text.

5. The computer-implemented method of claim 1 , further comprising opting to provide either the indication that the first entity attribute is similar to the second entity attribute or different than the second entity attribute based on one or more compare or contrast signals contained in a question posed about the segment of text.

6. The computer-implemented method of claim 1 , wherein identifying the first entity attribute of the first entity includes identifying a pronoun associated with the first entity attribute and co-reference resolving the first entity with the pronoun.

7. The computer-implemented method of claim 1 , wherein the providing the indication that the first entity attribute is either similar to or different than the second entity is based on one or more words that signify a comparison or contrast to be made between two components.

8. The computer-implemented method of claim 1 , wherein the providing the indication that the first entity attribute is either similar to or different than the second entity attribute comprises providing an indication that the first entity is similar to or different than the second entity based on a similarity or difference, respectively, between the first entity attribute and the second entity attribute.

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

receiving a question that includes information, the information including the first entity and the second entity;

matching the information to the first entity and the second entity; and

wherein the providing the indication that the first entity attribute is either similar to or different than the second entity attribute is in response to the matching.

10. The computer-implemented method of claim 1 , wherein the first and second entities are first and second products, the first and second entity attributes are first and second product attributes, and the method further comprises providing a summary of product features of the first and second products, the summary including an indication of the first and second product features.

11. A computer-implemented method, comprising:

identifying, by a computing system based at least in part on a segment of text a first entity and a second entity that is distinct from the first entity;

determining, based on a portion of the segment of text that directly describes a relationship between the first entity and a first entity attribute of the first entity and indirectly describes a relationship between the second entity and a second entity attribute of the second entity, the second entity attribute;

determining, by the computing system based at least in part on the segment of text, that the first entity attribute associated with the first entity in the segment of text and the second entity attribute associated with the second entity in the segment of text share an attribute class;

in response to the determining that the first entity attribute and the second entity attribute share an attribute class, providing, by the computing system, an indication that, in the segment of text, the first entity attribute is either similar to or different than the second entity attribute; and

generating a textual rewrite that directly describes a relationship between the second entity and the second entity attribute.

12. The computer-implemented method of claim 11 , wherein providing the indication that the first entity attribute is either similar to or different than the second entity attribute comprises providing the textual rewrite.

13. A system including memory and one or more processors operable to execute instructions stored in the memory, the memory comprising instructions to:

identify, in a segment of text, a subject and an action performed by the subject;

determine a likelihood that the action is performable by a class of subjects with which the subject is associated, the likelihood based at least in part on a plurality of reference subjects and associated reference actions found in a corpus of textual documents; and

provide an indication of the likelihood.

14. The system of claim 13 , wherein the system further comprises instructions to provide an indication that performance of the action by the subject is plausible or implausible based on the likelihood.

15. The system of claim 13 , wherein the subject is a non-human subject, and the system further comprises instructions to provide an indication that the subject has been personified in the segment of text.

16. The system of claim 13 , wherein the identifying comprises identifying the subject and action based at least in part on annotations associated with the segment of text.

17. The system of claim 13 , wherein the system further comprises instructions to determine the likelihood based on a question posed about the segment of text.

18. The system of claim 13 , wherein the system further comprises instructions to determine the likelihood based at least in part statistics derived from the plurality of reference subjects and associated reference actions found in the corpus of textual documents.

19. The system of claim 18 , wherein the system further comprises instructions to restrict the corpus to a textual documents that contain below a predetermined threshold of personified entities.

20. A non-transitory computer-readable medium comprising instructions that, in response to execution of the instructions by a computing system, cause the computing system to perform operations comprising:

identifying, based at least in part on a segment of text, a first entity and a second entity that is distinct from the first entity;

determining, based at least in part on the segment of text, that a first entity attribute associated with the first entity in the segment of text and a second entity attribute associated with the second entity in the segment of text share an attribute class;

in response to the determining, providing an indication that, in the segment of text, the first entity attribute is either similar to or different than the second entity attribute;

identifying, in the segment of text, a subject and an action performed by the subject;

determining a likelihood that the action is performable by a class of subjects with which the subject is associated, the likelihood based at least in part on a plurality of reference subjects and associated reference actions found in a corpus of textual documents; and

providing an indication of the likelihood.