Systems for dynamically generating and presenting narrative content

• US 10,176,170 B2
• Filed: 01/03/2017
• Issued: 01/08/2019
• Est. Priority Date: 04/06/2010
• Status: Active Grant

• Current Assignee: STATS LLC (Stats Holdings LLC)
• Original Assignee: Automated Insights, Inc. (Stats Holdings LLC)
• Inventors: Allen, Robert C.
• Primary Examiner(s): Godbold, Douglas
• Application Number:
  US15/396,816
• Publication Number:
  US 20170235723A1
• Time in Patent Office: 735 Days
• Field of Search: None
• CPC Class Codes: G06F 40/10 Text processing natural language analysis G06F40/20; semantic analysis G06F40/30; processing or translation of natural language G06F40/40
• CPC Class Codes: G06F 40/186 Templates
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/56 Natural language generation
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management

Abstract

In some embodiments, a non-transitory processor-readable medium stores code representing instructions that when executed cause a processor to select a narrative content template based at least in part on a predetermined content type associated with a real-world and/or virtual event. The code further represents instructions that when executed cause the processor to select a narrative tone type. The code further represents instructions that when executed cause the processor to, for each phrase included in an ordered set of phrases associated with the narrative content template, select, based at least in part on the narrative tone type, a phrase variation from a set of phrase variations associated with that phrase, and define, based on the selected phrase variation and at least one datum from a set of data, a narrative content portion associated with the real-world event. The code further represents instructions that when executed cause the processor to output, at a display, the narrative content portion.

Claims

1. A non-transitory processor-readable medium storing code representing instructions that when executed cause a processor to:

receive an event identifier (ID) associated with a real-world event;

receive a first participant ID associated with the event ID;

receive a statistic value associated with the first participant ID;

analyze the statistic in relation to statistics obtained from a statistics database to identify a statistical item of interest;

determine a narrative tone;

define, based on a content template, a content portion including;

a name of a first participant associated with the first participant ID;

the statistical item of interest; and

narrative text associated with the event, the text based at least in part on a plurality of text elements associated with the content template, each text element from the plurality of text elements being associated with the narrative tone; and

send the content portion to a physical output device for presentation;

wherein sending the content portion to a physical device comprises sending the content portion to a voice synthesizer for conversion to voice data for audible presentation.

2. The non-transitory processor-readable medium of claim 1, wherein the narrative tone is determined based at least in part on the statistical item of interest.

3. The non-transitory processor-readable medium of claim 2 wherein the narrative tone is further determined based at least in part on at least one of:

a recent performance of the first participant;

a ranking of the first participant; and

a physical location of the real-world event.

4. The non-transitory processor-readable medium of claim 1, further storing code representing instructions that when executed cause a processor to:

select a voice from a plurality of different voices; and

instruct the voice synthesizer to convert the narrative content portion to synthesized voice data based on the selected voice.

5. The non-transitory processor-readable medium of claim 1, wherein each text element from the plurality of text elements is one of:

a phrase;

a sentence;

or

a paragraph.

6. The non-transitory processor-readable medium of claim 1, wherein the statistical item of interest is a trend that occurred as a result of the real-world event.

7. An apparatus for automated natural language generation, comprising:

a statistics analysis module, the statistics analysis module for receiving statistics for a real-world event and analyzing the statistics by running one or more algorithms on at least the statistics to determine one or more particular items of interest related to the real-world event;

a processor to be operatively coupled to a memory and to the statistics analysis module, for executing content generation steps comprising;

receiving an instruction to define a content portion based on the real-world event;

receiving, from the statistics analysis module, a determined item of interest related to the real-world event;

selecting, based at least in part on the determined item of interest, a narrative tone type;

selecting, based at least in part on the determined item of interest, a narrative template that includes a set of phrases;

selecting, based on the narrative tone type, a phrase variation from a set of phrase variations associated with a first phrase from the set of phrases to define a first selected phrase;

selecting, based on the narrative tone type, a phrase variation from a set of phrase variations associated with a second phrase from the set of phrases to define a second selected phrase;

sending a signal indicative of the narrative content portion that includes the first selected phrase and the second selected phrase; and

sending a signal to an output device for outputting the narrative content portion; and

a voice synthesizer for;

receiving the signal indicative of the narrative content portion;

converting the narrative content portion to voice data; and

outputting the voice data audibly.

8. The apparatus of claim 7, the voice synthesizer further for selecting a voice from a plurality of different voices, and for converting the narrative content portion to voice data based on the selected voice.

9. The apparatus of claim 7, the statistics analysis module for acting in substantially real-time to determine the item of interest during the performance of the real-world event.

10. The apparatus of claim 7, wherein the real-world event is a sporting contest.

11. The apparatus of claim 7, wherein the item of interest is a trend that occurred as a result of the real-world event.

12. A method for providing automated natural language generation, comprising:

receiving a digital input representing a real-world event;

receiving an instruction to define a narrative content portion based on the real-world event;

receiving, from a database, a plurality of data associated with the event, including statistics for the real-world event;

running one or more algorithms on at least the received statistics to determine one or more particular items of interest related to the real-world event;

selecting, based at least in part on the determined item of interest, a narrative tone type;

selecting, based at least in part on the determined item of interest, a narrative content template including a plurality of variable identifiers;

defining the narrative content portion based at least in part on a combination of;

a text portion of the narrative content template, the text portion being based on the narrative tone type; and

for each variable identifier from the plurality of variable identifiers, a datum from the plurality of data associated with that variable identifier; and

sending a signal such that the narrative content portion is synthesized as output on an output device;

wherein the output is synthesized vocal output.

13. The method of claim 12 further comprising:

selecting a voice from a plurality of different voices, and synthesizing the narrative content portion as vocal output based on the selected voice.

14. The method of claim 12, wherein determining the item of interest occurs in substantially real-time during the performance of the real-world event.

15. The method of claim 12, wherein determining the item of interest further comprises:

identifying a trend that occurred as a result of the performance of the real-world event; and

wherein the synthesized output includes synthesized output of the identified trend.

16. The method of claim 12, wherein the real-world event is a sporting contest.

17. The method of claim 16, wherein the item of interest is based on a team'"'"'s historical winning record against a betting point spread.