CPC G06N 5/02 (2013.01) [G06F 40/56 (2020.01)] | 20 Claims |
1. A method for applying artificial intelligence technology to automatically generate narrative stories about a data set, the method comprising:
executing a database query to retrieve from a subject data database system data that represents a set of subjects about which the data set is to be evaluated to determine whether to trigger generation of a narrative story about the data set that pertains to a subject from among the set of subjects, wherein a plurality of the subjects are associated with different story angles, wherein each of a plurality of the story angles is associated with a characterization of the data set and comprises a plurality of applicability conditions for determining whether the characterization fits the data set;
evaluating the data set against the different story angles via a processor by (1) testing the applicability conditions of the different story angles to determine whether any of the characterizations corresponding to the different story angles fit the data set, wherein a subject from among the set of subjects that corresponds to a characterization determined to fit the data set serves as a relevant subject, and (2) generating an evaluation indicator based on the testing step, wherein the evaluation indicator indicates whether a narrative story about the relevant subject is to be generated;
determining a data model representation of the narrative story based on the data set, the data representation including one or more data elements and a plurality of procedures;
generating the narrative story about the relevant subject upon determining that the evaluation indicator indicates that the narrative story about the relative subject is to be generated, wherein generating the story comprises recursively traversing the representation to apply a plurality of phrasal generation routines associated with data types, wherein the rendered narrative story expresses the characterization of the data set that was determined to fit the data set; and
repeating the executing, evaluating, and generating steps over time as the data set is updated with new data.
|