Method and apparatus for generating a linguistic representation of raw input data

• US 10,776,561 B2
• Filed: 01/15/2013
• Issued: 09/15/2020
• Est. Priority Date: 01/15/2013
• Status: Active Grant

• Current Assignee: Arria Data2text Limited
• Original Assignee: Arria Data2text Limited
• Inventors: Sripada, Gowri Somayajulu
• Primary Examiner(s): Paula, Cesar B
• Assistant Examiner(s): Nguyen, Luu-Phuong T
• Application Number:
  US14/760,848
• Publication Number:
  US 20150363364A1
• Time in Patent Office: 2,800 Days
• Field of Search: 715243
• CPC Class Codes: G06F 16/211 Schema design and management
• CPC Class Codes: G06F 16/258 Data format conversion from or to a database
• CPC Class Codes: G06F 16/95 Retrieval from the web
• CPC Class Codes: G06F 40/103 Formatting, i.e. changing of presentation of documents automatic justification G06F40/189; automatic line break hyphenation G06F40/191
• CPC Class Codes: G06F 40/106 Display of layout of documents; Previewing
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30

Abstract

Methods, apparatuses, and computer program products are described herein that are configured to be embodied as and/or performed by a document planner. In some example embodiments, a method is provided that comprises selecting a schema based on one or more messages available in a message store and using the selected schema and one or more messages available in the message store to generate a document plan. The schema of this embodiment may be defined by a specification containing one or more queries for selecting one or more messages, one or more messages, and/or one or more predefined phrases to instantiate a document plan. The method of this embodiment may also include applying an optimization specification to the document plan to generate an optimized document plan.

Claims

1. A computer-implemented method for generating an output text by transforming received raw input data into a format that can be linguistically expressed in the output text comprising:

selecting, using a processor, a schema from a schema store, wherein the schema is selected based on one or more messages in a message store, each of the one or more messages being a language independent data structure, each message corresponding to and selected in response to an existence of at least one fact about the received raw input data, the received raw input data expressed in a non-linguistic format;

generating, using the processor, a document plan that is instantiated with the one or more messages, wherein the document plan is instantiated according to the schema that comprises a query set, at least one query in the query set configured for selecting any number of the one or more messages in the message store;

altering, using the processor, the document plan according to an optimization specification, wherein the optimization specification comprises a set of one or more rules, the rules comprising at least one rule for arranging the one or more messages of the document plan based on an optimization function; and

generating, using the processor, an output text for display on a user interface based on the altered document plan, the output text being a linguistic representation of the raw input data.

2. A method according to claim 1, wherein generating the document plan further comprises the selected schema selecting, using the processor, a second schema to also be used in the generating of the document plan.

3. A method according to claim 1, wherein the schema further comprises one or more of pre-defined text and queries to be executed against the message store for use in generating the document plan.

4. A method according to claim 3, wherein an alternate query may be specified if an initial query executed against the message store does not return a result including one or more messages.

5. A method according to claim 1, wherein the document plan further comprises nodes having parameters specifying one or more of document structure, rhetorical structure, message ordering, document plan optimization, and microplanner parameters.

6. A method according to claim 1, wherein the document plan further comprises nodes having microplanner parameters.

7. A method according to claim 1, wherein the optimization specification further comprises rules for at least one or more of text size, text fluency, avoiding repetition, determining paragraph breaks, ordering messages, ensuring narrative coherence, maintaining discourse focus, and narration development.

8. A method according to claim 7, wherein the optimization specification rules further comprise triggering conditions defined over at least one or more of message properties, message types, and relationships of messages.

9. A method according to claim 1, wherein the optimization specification further specifies one or more of sequencing patterns of messages and aggregation of messages.

10. A method according to claim 1, wherein the one or more rules are configured for comparing a value for at least one message property of the one or more messages in the message store to arrange the one or more messages for maintaining discourse focus.

11. An apparatus that is configured to generate an output text by transforming received raw input data into a format that can be linguistically expressed in the output text, 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;

select, using the at least one processor, a schema from a schema store, wherein the schema is selected based on one or more messages in a message store, each of the one or more messages being a language independent data structure, each message corresponding to and selected in response to an existence of at least one fact about the received raw input data, the received raw input data expressed in a non-linguistic format;

generate, using the at least one processor, a document plan that is instantiated with the one or more messages, wherein the document plan is instantiated according to the schema that comprises a query set, at least one query in the query set configured for selecting any number of the one or more messages in the message store;

alter, using the at least one processor, the document plan in accordance with an optimization specification, wherein the optimization specification comprises a set of one or more rules, the rules comprising at least one rule for arranging the one or more messages of the document plan based on an optimization function; and

generate, using the at least one processor, an output text for display on a user interface based on the altered document plan, the output text being a linguistic representation of the raw input data.

12. An apparatus according to claim 11, wherein generating the document plan further comprises the selected schema selecting a second schema to also be used in the generating of the document plan.

13. An apparatus according to claim 11, wherein the schema further comprises one or more of pre-defined text and queries to be executed against the message store for use in generating the document plan.

14. An apparatus according to claim 13, wherein an alternate query may be specified if an initial query executed against the message store does not return a result including one or more messages.

15. An apparatus according to claim 11, wherein the document plan further comprises nodes having parameters specifying one or more of document structure, rhetorical structure, message ordering, document plan optimization, and microplanner parameters.

16. An apparatus according to claim 11, wherein the optimization specification comprises rules for at least one or more of text size, text fluency, avoiding repetition, determining paragraph breaks, ordering messages, ensuring narrative coherence, maintaining discourse focus, and narration development.

17. An apparatus according to claim 16, wherein the optimization specification rules further comprise triggering conditions defined over at least one or more of message properties, message types, and relationships of messages.

18. An apparatus according to claim 11, wherein the optimization specification specifies one or more of sequencing patterns of messages and aggregation of messages.

19. An apparatus according to claim 11, wherein the one or more rules are configured to compare a value for at least one message property of the one or more messages in the message store to arrange the one or more messages for maintaining discourse focus.

20. A computer program product that is configured to generate an output text by transforming received raw input data into a format that can be linguistically expressed in the output text, 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 having at least one processor cause the apparatus at least to;

select, using the at least one processor, a schema from a schema store, wherein the schema is selected based on one or more messages in a message store, each of the one or more messages being a language independent data structure, each message corresponding to and selected in response to an existence of at least one fact about the received raw input data, the received raw input data expressed in a non-linguistic format;

generate, using the at least one processor, a document plan that is instantiated with the one or more messages, wherein the document plan is instantiated according to the schema that comprises a query set, at least one query in the query set configured for selecting any number of the one or more messages in the message store;

alter, using the at least one processor, the document plan in accordance with an optimization specification, wherein the optimization specification comprises a set of one or more rules, the rules comprising at least one rule for arranging the one or more messages of the document plan based on an optimization function; and

generate, using the at least one processor, an output text for display on a user interface based on the altered document plan the output text being a linguistic representation of the raw input data.