Applied artificial intelligence technology for natural language generation using a story graph and configurable structurer code
Natural language generation technology is disclosed that applies artificial intelligence to structured data to determine content for expression in natural language narratives that describe the structured data. A story graph is employed, where the story graph comprises a plurality of items of content to be expressed in a natural language narrative. A processor executes structurer code based on a plurality of operating rules and/or parameters that control how the structurer code organizes the items of content into a story outline from which the natural language narrative is generated, wherein the operating rules and/or parameters are configurable to change strategies for organizing the items of content into the story outline.
1 . A natural language generation (NLG) system that applies artificial intelligence to organize content to be expressed in a natural language narrative that describes structured data, the system comprising:
a memory configured to store a story graph comprising a plurality of nodes corresponding to items of content to be expressed in a natural language narrative, a node of the plurality of nodes characterizing a quantitative result determined by executing parameterized analytics mapping one or more variables to the structured data; and
a processor configured to:
execute structurer code based on a plurality of operating rules or parameters that control how the structurer code organizes the items of content into a story outline from which the natural language narrative is generated, the operating rules or parameters being configured to change strategies for organizing the items of content into the story outline, wherein the operating rules or parameters are configured based on training data associated with a user to whom the natural language narrative is to be presented, wherein the training data includes user feedback, and
execute a natural language generation artificial intelligence model to determine the natural language narrative based on the story outline and characterizing the quantitative result in natural language.
2 . The system of claim 1 wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on importance values assigned to the items of content.
3 . The system of claim 1 wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on interestingness values assigned to the items of content.
4 . The system of claim 1 wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on characterizations assigned to the items of content.
5 . The system of claim 4 wherein the characterizations comprise positive characterizations and/or a negative characterizations.
6 . The system of claim 1 wherein the items of content are generated from a plurality of nodes of a graph data structure.
7 . The system of claim 6 wherein the nodes are related in the graph data structure via a plurality of links, wherein the links include metadata that describe an aspect of a relationship between the linked nodes, and wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on the links metadata.
8 . The system of claim 6 wherein the graph data structure is adjustable to add one or more additional nodes to the graph data structure.
9 . The system of claim 6 wherein the graph data structure is adjustable to modify one or more of the nodes.
10 . The system of claim 6 wherein the graph data structure is parameterized based on the structured data.
11 . The system of claim 6 wherein the graph data structure comprises an authoring graph.
12 . The system of claim 11 wherein the authoring graph is parameterized based on the structured data to define a knowledge graph, wherein the processor operates on the knowledge graph to determine the story graph.
13 . The system of claim 1 wherein the processor is configured to generate a plurality of the natural language narratives based on a plurality of story graphs in an interactive mode based on conversational inputs from users.
14 . The system of claim 1 wherein the structurer code determines, via the story outline, an order in which the items of content are to be expressed in the natural language narrative.
15 . The system of claim 1 wherein the processor comprises a plurality of processors.
16 . A natural language generation (NLG) method that applies artificial intelligence to organize content to be expressed in a natural language narrative that describes structured data, the system comprising:
a processor accessing a story graph in memory, the story graph comprising a plurality nodes corresponding to items of content to be expressed in a natural language narrative, a node of the plurality of nodes characterizing a quantitative result determined by executing parameterized analytics mapping one or more variables to the structured data; and
the processor executing structurer code based on a plurality of operating rules and/or parameters that control how the structurer code organizes the items of content into a story outline from which the natural language narrative is generated, the operating rules or parameters being configured to change strategies for organizing the items of content into the story outline, the processing being configured to execute a natural language generation artificial intelligence model to determine the natural language narrative based on the story outline and characterizing the quantitative result in natural language, wherein the operating rules or parameters are configured based on training data associated with a user to whom the natural language narrative is to be presented, wherein the training data includes user feedback.
17 . An article of manufacture for natural language generation (NLG) that applies artificial intelligence to structured data to determine content to be expressed in natural language narratives that describe the structured data, the article of manufacture comprising:
machine-readable code that is resident on a non-transitory computer-readable storage medium, wherein the code is executable by a processor to cause the processor to:
access a story graph in memory, the story graph comprising a plurality of nodes corresponding to items of content to be expressed in a natural language narrative, a node of the plurality of nodes characterizing a quantitative result determined by executing parameterized analytics mapping one or more variables to the structured data;
execute structurer code based on a plurality of operating rules or parameters that control how the structurer code organizes the items of content into a story outline from which the natural language narrative is generated, the operating rules or parameters being configured to change strategies for organizing the items of content into the story outline, wherein the operating rules or parameters are configured based on training data associated with a user to whom the natural language narrative is to be presented, wherein the training data includes user feedback; and
execute a natural language generation artificial intelligence model to determine the natural language narrative based on the story outline and characterizing the quantitative result in natural language.
18 . The article of manufacture of claim 17 wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on (1) importance values assigned to the items of content and/or (2) interestingness values assigned to the items of content.
19 . The article of manufacture of claim 17 wherein the operating rules or parameters are configurable to organize the items of content into the story outline based on characterizations assigned to the items of content.