Automatic generation of narratives from data using communication goals and narrative analytics
The exemplary embodiments described herein are related to techniques for automatically generating narratives about data based on communication goal data structures that are associated with configurable content blocks. The use of such communication goal data structures facilitates modes of operation whereby narratives can be generated in real-time and/or interactive manners.
1. A natural language generation (NLG) system for interacting with a data set and artificial intelligence to generate natural language narrative content about the data set in response to natural language inputs, the system comprising:
a memory configured to store (1) a data set of structured data, (2) a plurality of data structures that explicitly represent a plurality of different communication goals, and (3) a plurality of narrative analytics; and
a processor configured to (1) process a natural language input, (2) select the data structure that explicitly represents a communication goal indicated by the processed natural language input, wherein the selected data structure is linked to a set of the narrative analytics, (3) parameterize the linked set of narrative analytics by defining a plurality of values for parameters of the linked set of narrative analytics based on the processed natural language input, (4) execute the parameterized set of narrative analytics with respect to the data set to determine content to be expressed in a natural language narrative that (i) is responsive to the processed natural language input and (ii) satisfies the communication goal indicated by the processed natural language input, and (5) generate the natural language narrative about the data set based on the determined content.
2. The system of claim 1 wherein the processor is configured to perform its operations in response to a plurality of natural language inputs to produce an interactive natural language dialog about the data set.
3. The system of claim 2 wherein the memory is further configured to store a plurality of data models;
wherein the selected data structure is linked to one of the data models; and
wherein the processor is further configured to (1) parameterize the linked data model by defining a plurality of values for parameters of the linked data model based on the processed natural language input and (2) execute the parameterized set of narrative analytics based on the parameterized data model with respect to the data set to determine the content to be expressed in the natural language narrative.
4. The system of claim 1 wherein the natural language input comprises a natural language query from a user.
5. The system of claim 1 wherein the memory is further configured to store a plurality of content block data structures, wherein each of a plurality of the content block data structures are linked to a set of the narrative analytics, wherein the data structures that explicitly represent the communication goals are linked to the content block data structures, and wherein the processor is further configured to determine the linked set of narrative analytics by identifying the set of narrative analytics that are linked to the content block data structure linked to the data structure that explicitly represents the communication goal indicated by the processed natural language input.
6. The system of claim 5 wherein a plurality of the content block data structures include a subject parameter that generalizes those content block data structures to pertain to a plurality of different subjects, and wherein the processor is further configured to parameterize the linked content block data structure to a specific subject by defining a value for the subject parameter to focus the content block data structure on the specific subject.
7. The system of claim 5 wherein a plurality of the content block data structures include a feature parameter that generalizes those content block data structures to pertain to a plurality of different features, and wherein the processor is further configured to parameterize the linked content block data structure by defining a value for the feature parameter to focus the parameterized content block data structure on a feature corresponding to the defined feature value.
8. The system of claim 1 wherein the narrative analytics specify a plurality of angles for testing against the data set to identify an angle that is deemed to accurately characterize the data set, each angle corresponding to a different characterization, and wherein the natural language narrative expresses the characterization corresponding to an angle deemed by the testing to accurately characterize the data set.
9. The system of claim 1 wherein the explicit representation of the indicated communication goal corresponds to a describe communication goal.
10. The system of claim 1 wherein the explicit representation of the indicated communication goal corresponds to an evaluate communication goal.
11. The system of claim 1 wherein the explicit representation of the indicated communication goal corresponds to an explain communication goal.
12. The system of claim 1 wherein the memory comprises a plurality of memories.
13. The system of claim 1 wherein the processor comprises a plurality of processors.
14. A natural language generation (NLG) method for interacting with a data set and artificial intelligence to generate natural language narrative content about the data set in response to natural language inputs, the method comprising:
a processor processing a natural language input;
the processor accessing a memory, wherein the memory stores (1) a data set of structured data, (2) a plurality of data structures that explicitly represent a plurality of different communication goals, and (3) a plurality of narrative analytics;
the processor selecting the data structure that explicitly represents a communication goal indicated by the processed natural language input, wherein the selected data structure is linked to a set of the narrative analytics;
the processor parameterizing the linked set of narrative analytics by defining a plurality of values for parameters of the linked set of narrative analytics based on the processed natural language input;
the processor executing the parameterized set of narrative analytics with respect to the data set to determine content to be expressed in a natural language narrative that (i) is responsive to the processed natural language input and (ii) satisfies the communication goal indicated by the processed natural language input; and
the processor generating the natural language narrative about the data set based on the determined content.
15. The method of claim 14 further comprising:
the processor performing its operations in response to a plurality of natural language inputs to produce an interactive natural language dialog about the data set.
16. The method of claim 14 wherein the memory comprises a plurality of memories.
17. The method of claim 14 wherein the processor comprises a plurality of processors.
18. A natural language generation (NLG) computer program product for interacting with a data set and artificial intelligence to generate natural language narrative content about the data set in response to natural language inputs, the computer program product comprising:
a non-transitory computer-readable storage medium configured to store (1) a data set of structured data, (2) a plurality of data structures that explicitly represent a plurality of different communication goals, and (3) a plurality of narrative analytics; and
a plurality of instructions that are resident on the non-transitory computer-readable storage medium, wherein the instructions are configured for execution by a processor to cause the processor to (1) process a natural language input, (2) select the data structure that explicitly represents a communication goal indicated by the processed natural language input, wherein the selected data structure is linked to a set of the narrative analytics, (3) parameterize the linked set of narrative analytics by defining a plurality of values for parameters of the linked set of narrative analytics based on the processed natural language input, (4) execute the parameterized set of narrative analytics with respect to the data set to determine content to be expressed in a natural language narrative that (i) is responsive to the processed natural language input and (ii) satisfies the communication goal indicated by the processed natural language input, and (5) generate the natural language narrative about the data set based on the determined content.
19. The computer program product of claim 18 wherein the instructions are configured for execution in response to a plurality of natural language inputs to produce an interactive natural language dialog about the data set.
20. The computer program product of claim 18 wherein the non-transitory computer-readable storage medium comprises a plurality of non-transitory computer-readable storage media.