Method and apparatus for aggregating with information generalization

• US 10,216,728 B2
• Filed: 02/01/2017
• Issued: 02/26/2019
• Est. Priority Date: 11/02/2012
• Status: Active Grant

• Current Assignee: Arria Data2text Limited
• Original Assignee: Arria Data2text Limited
• Inventors: Bradshaw, William Anthony, Reiter, Ehud Baruch
• Primary Examiner(s): Godbold, Douglas
• Application Number:
  US15/421,925
• Publication Number:
  US 20170364511A1
• Time in Patent Office: 755 Days
• Field of Search: None
• CPC Class Codes: G06F 40/205 Parsing
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06F 40/56 Natural language generation

Abstract

Methods, apparatuses, and computer program products are described herein that are configured to perform aggregation of phrase specifications. In some example embodiments, a method is provided that comprises identifying two or more generalized phrase specifications. In some example embodiments, the two or more generalized phrase specifications contain at least one aggregatable constituent. The method of this embodiment may also include generating an aggregated phrase specification from the two or more generalized phrase specifications. In some example embodiments, the aggregated phrase specification comprises a combined noun phrase generated from the aggregatable constituents and one or more additional constituents based on a determined level of generalization.

Claims

1. A method for generating language by transforming raw input data that is at least partially expressed in a non-linguistic format into a format that can be expressed linguistically in a textual output, the method comprising:

generating a plurality of phrase specifications from the raw input data;

identifying a domain model based on the plurality of phrase specifications, wherein the domain model includes at least one domain rule;

identifying one or more phrase specifications from the plurality of phrase specifications, wherein each of the one or more phrase specifications contains at least one aggregatable constituent;

generating one or more generalized phrase specifications from the one or more phrase specifications based at least in part on the at least one domain rule; and

generating the textual output, based at least in part on the one or more generalized phrase specifications, such that is it displayable on a user interface.

2. A method according to claim 1, further comprising:

identifying two or more generalized phrase specifications, wherein each of the two or more generalized phrase specifications contains the at least one aggregatable constituent;

generating an aggregated phrase specification from the two or more generalized phrase specifications based at least in part on the at least one domain rule, wherein the aggregated phrase specification comprises at least one of a combined noun phrase generated from the at least one aggregatable constituent and one or more additional constituents based on a determined level of generalization; and

generating the textual output, based at least in part on the aggregated phrase specification, such that it is displayable on a user interface.

3. A method according to claim 2, further comprising:

determining one or more constituents that are removable in one or more phrase specifications based at least in part on the at least one domain rule; and

removing the one or more constituents that are removable, such that the one or more phrase specifications are one or more generalized phrase specifications.

4. A method according to claim 1, further comprising:

determining one or more constituents that are generalizable in one or more phrase specifications;

determining a generalized constituent for at least one of the one or more generalizable constituents based at least in part on the at least one domain rule; and

replacing the one or more generalized constituents with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

5. A method according to claim 4, wherein the generalized constituent is a most generalized constituent in a predefined constituent listing.

6. A method according to claim 3, further comprising:

removing all of the one or more constituents from the one or more phrase specifications that are removable; and

replacing all of the one or more generalizable constituents from the one or more phrase specifications with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

7. A method according to claim 1, wherein the two or more generalized phrase specifications are identified as aggregatable in an instance in which each of the two or more generalized phrase specifications are identical but for the at least one aggregatable constituent in each of the two or more generalized phrase specifications.

8. A method according to claim 1, further comprising:

generating the combined noun phrase using the at least one aggregatable constituent in the two or more generalized phrase specifications.

9. A method according to claim 1, further comprising:

determining that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is removed; and

populating the aggregated phrase specification with at least one of a generalized constituent or a removed constituent.

10. A method according to claim 1, further comprising:

determining that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is less generalized than a generalized constituent; and

populating the aggregated phrase specification with the constituent that is less generalized than the generalized constituent based on a predefined constituent listing.

11. An apparatus for generating language by transforming raw input data that is at least partially expressed in a non-linguistic format into a format that can be expressed linguistically in a textual output, the apparatus comprising:

at least one processor; and

at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least;

generate a plurality of phrase specifications from the raw input data;

identify a domain model based on the plurality of phrase specifications, wherein the domain model includes at least one domain rule;

identify one or more phrase specifications from the plurality of phrase specifications, wherein each of the one or more phrase specifications contains at least one aggregatable constituent;

generate one or more generalized phrase specifications from the one or more phrase specifications based at least in part on the at least one domain rule; and

generate the textual output, based at least in part on the one or more generalized phrase specifications, such that is it displayable on a user interface.

12. An apparatus according to claim 11, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

identify two or more generalized phrase specifications, wherein each of the two or more generalized phrase specifications contains the at least one aggregatable constituent;

generate an aggregated phrase specification from the two or more generalized phrase specifications based at least in part on the at least one domain rule, wherein the aggregated phrase specification comprises at least one of a combined noun phrase generated from the at least one aggregatable constituent and one or more additional constituents based on a determined level of generalization; and

generate the textual output, based at least in part on the aggregated phrase specification, such that it is displayable on a user interface.

13. An apparatus according to claim 12, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

determine one or more constituents that are removable in one or more phrase specifications based at least in part on the at least one domain rule; and

remove the one or more constituents that are removable, such that the one or more phrase specifications are one or more generalized phrase specifications.

14. An apparatus according to claim 11, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

determine one or more constituents that are generalizable in one or more phrase specifications;

determine a generalized constituent for at least one of the one or more generalizable constituents based at least in part on the at least one domain rule; and

replace the one or more generalized constituents with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

15. An apparatus according to claim 14, wherein the generalized constituent is a most generalized constituent in a predefined constituent listing.

16. An apparatus according to claim 13, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

remove all of the one or more constituents from the one or more phrase specifications that are removable; and

replace all of the one or more generalizable constituents from the one or more phrase specifications with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

17. An apparatus according to claim 11, wherein the two or more generalized phrase specifications are identified as aggregatable in an instance in which each of the two or more generalized phrase specifications are identical but for the at least one aggregatable constituent in each of the two or more generalized phrase specifications.

18. An apparatus according to claim 11, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

generate the combined noun phrase using the at least one aggregatable constituent in the two or more generalized phrase specifications.

19. An apparatus according to claim 11, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

determine that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is removed; and

populate the aggregated phrase specification with at least one of a generalized constituent or a removed constituent.

20. An apparatus according to claim 11, wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:

determine that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is less generalized than a generalized constituent; and

populate the aggregated phrase specification with the constituent that is less generalized than the generalized constituent based on a predefined constituent listing.

21. A computer program product for generating language by transforming raw input data that is at least partially expressed in a non-linguistic format into a format that can be expressed linguistically in a textual output, the computer program product comprising:

at least one computer readable non-transitory memory medium having program code instructions stored thereon, the program code instructions, which when executed by an apparatus, cause the apparatus at least to;

generate a plurality of phrase specifications from the raw input data;

identify a domain model based on the plurality of phrase specifications, wherein the domain model includes at least one domain rule;

identify one or more phrase specifications from the plurality of phrase specifications, wherein each of the one or more phrase specifications contains at least one aggregatable constituent;

generate one or more generalized phrase specifications from the one or more phrase specifications based at least in part on the at least one domain rule; and

generate the textual output, based at least in part on the one or more generalized phrase specifications, such that is it displayable on a user interface.

22. A computer program product according to claim 21 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

identify two or more generalized phrase specifications, wherein each of the two or more generalized phrase specifications contains the at least one aggregatable constituent;

generate an aggregated phrase specification from the two or more generalized phrase specifications based at least in part on the at least one domain rule, wherein the aggregated phrase specification comprises at least one of a combined noun phrase generated from the at least one aggregatable constituent and one or more additional constituents based on a determined level of generalization; and

generate the textual output, based at least in part on the aggregated phrase specification, such that it is displayable on a user interface.

23. A computer program product according to claim 22 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

determine one or more constituents that are removable in one or more phrase specifications based at least in part on the at least one domain rule; and

remove the one or more constituents that are removable, such that the one or more phrase specifications are one or more generalized phrase specifications.

24. A computer program product according to claim 21 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

determine one or more constituents that are generalizable in one or more phrase specifications;

determine a generalized constituent for at least one of the one or more generalizable constituents based at least in part on the at least one domain rule; and

replace the one or more generalized constituents with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

25. A computer program product according to claim 24, wherein the generalized constituent is a most generalized constituent in a predefined constituent listing.

26. A computer program product according to claim 23 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

remove all of the one or more constituents from the one or more phrase specifications that are removable; and

replace all of the one or more generalizable constituents from the one or more phrase specifications with the determined generalized constituent, such that the one or more phrase specifications are one or more generalized phrase specifications.

27. A computer program product according to claim 21, wherein the two or more generalized phrase specifications are identified as aggregatable in an instance in which each of the two or more generalized phrase specifications are identical but for the at least one aggregatable constituent in each of the two or more generalized phrase specifications.

28. A computer program product according to claim 21 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

generate the combined noun phrase using the at least one aggregatable constituent in the two or more generalized phrase specifications.

29. A computer program product according to claim 21 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

determine that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is removed; and

populate the aggregated phrase specification with at least one of a generalized constituent or a removed constituent.

30. A computer program product according any to claim 21 further comprises program code instructions, which when executed by an apparatus, further cause the apparatus at least to:

determine that the two or more generalized phrase specifications are still identified as aggregatable with a constituent that is less generalized than a generalized constituent; and

populate the aggregated phrase specification with the constituent that is less generalized than the generalized constituent based on a predefined constituent listing.