Methods, apparatuses and computer program products for providing a conversational data-to-text system

Patent metadata

Field	Value
Publication number	US-12072874-B2
Application number	US-202117462988-A
Country	US
Kind code	B2
Filing date	Aug 31, 2021
Priority date	Aug 31, 2020
Publication date	Aug 27, 2024
Grant date	Aug 27, 2024

Abstract

Methods, apparatuses and computer program products for providing a conversational data-to-text system are described herein. An example method may include receiving a first natural language query from a client device; generating a first analytic operation instruction associated with a multi-dimensional dataset based at least in part on the first natural language query; determining a first multi-dimensional data object based at least in part on the first analytic operation instruction and the multi-dimensional dataset; generating a first natural language response to the first natural language query based at least in part on the first multi-dimensional data object; and transmitting the first natural language response to the client device.

First claim

The invention claimed is:

1. An apparatus comprising at least one processor and at least one non-transitory memory comprising program code, the at least one non-transitory memory and the program code configured to, the at least one processor, cause the apparatus to at least:
receive a first natural language query from a client device;
generate, based at least in part on the first natural language query, one or more semantic frames;
retrieve, based at least in part on the one or more semantic frames, a multi-dimensional dataset comprising one or more multi-dimensional data objects each placed in a feature space of a plurality of feature spaces, wherein each feature space comprises one or more dimensions corresponding to categorical data;
generate, based at least in part on the one or more semantic frames, an expected structure of a resulting multi-dimensional data object and a first analytic operation query to be performed on the resulting multi-dimensional data object, wherein the first analytic operation query defines at least one analytic operation type of a plurality of analytic operation types;
generate, based at least in part on executing the first analytic operation query on the resulting multi-dimensional data object, insights associated with data represented by the resulting multi-dimensional data object;
generate a first natural language response to the first natural language query based at least in part on the insights;
convert the first natural language response from text to audio output; and
transmit the audio output of the first natural language response to the client device.

2. The apparatus of claim 1 , wherein the first analytic operation query further defines at least one query parameter.

3. The apparatus of claim 2 , wherein the at least one query parameter corresponds to a dimension instance in a feature space of the plurality of feature spaces associated with the multi-dimensional dataset.

4. The apparatus of claim 2 , wherein the at least one analytic operation type comprises one or more of a filtering operation, a grouping operation, or a variance operation.

5. The apparatus of claim 3 , wherein the feature space comprises a corresponding measure representing numerical data that a given multi-dimensional data object represents.

6. The apparatus of claim 1 , wherein the at least one non-transitory memory and the program code configured to, with the at least one processor, cause the apparatus to further:
select a narrative function script based at least in part on the first natural language query; and
generate the first natural language response further based at least in part on the narrative function script.

7. The apparatus of claim 1 , wherein, subsequent to transmitting the first natural language response to the client device, the at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to further:
update contextual data stored in a discourse model based at least in part on the first natural language query and the first natural language response;
receive a second natural language query from the client device;
generate, based at least in part on the second natural language query and the contextual data, a second analytic operation query associated with the multi-dimensional dataset;
generate, based at least in part on the multi-dimensional dataset and executing the second analytic operation query, a second multi-dimensional data object;
generate a second natural language response to the second natural language query based at least in part on the second multi-dimensional data object; and
transmit the second natural language response to the client device.

8. The apparatus of claim 7 , wherein, when generating the second analytic operation query, the at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to further:
generate at least one inferred query parameter based at least in part on the second natural language query and the contextual data.

9. The apparatus of claim 8 , wherein, when generating the second multi-dimensional data object, the at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to further:
generate the second multi-dimensional data object based at least in part on the at least one inferred query parameter.

10. A computer-implemented method comprising:
receiving a first natural language query from a client device;
generating, based at least in part on the first natural language query, one or more semantic frames;
retrieving, based at least in part on the one or more semantic frames, a multi-dimensional dataset comprising one or more multi-dimensional data objects each placed in a feature space of a plurality of feature spaces, wherein each feature space comprise one or more dimensions corresponding to categorical data;
generating, based at least in part on the one or more semantic frames, an expected structure of a resulting multi-dimensional data object and a first analytic operation query to be performed on the resulting multi-dimensional data object, wherein the first analytic operation query defines at least one analytic operation type of a plurality of analytic operation types;
generating, based at least in part on executing the first analytic operation query on the resulting multi-dimensional data object, insights associated with data represented by the resulting multi-dimensional data object;
generating a first natural language response to the first natural language query based at least in part on the insights;
converting the first natural language response from text to audio output; and
transmitting the audio output of the first natural language response to the client device.

11. The computer-implemented method of claim 10 , wherein the first analytic operation query further defines at least one query parameter.

12. The computer-implemented method of claim 11 , wherein the at least one query parameter corresponds to a dimension instance in a feature space of the plurality of feature spaces associated with the multi-dimensional dataset.

13. The computer-implemented method of claim 11 , wherein the at least one analytic operation type comprises one or more of a filtering operation, a grouping operation, or a variance operation.

14. The computer-implemented method of claim 11 , wherein the feature space comprises a corresponding measure representing numerical data that a given multi-dimensional data object represents.

15. The computer-implemented method of claim 10 , further comprising:
selecting a narrative function script based at least in part on the first natural language query; and
generating the first natural language response further based at least in part on the narrative function script.

16. The computer-implemented method of claim 10 , wherein, subsequent to transmitting the first natural language response to the client device, the computer-implemented method further comprises:
updating contextual data stored in a discourse model based at least in part on the first natural language query and the first natural language response;
receiving a second natural language query from the client device;
generating, based at least in part on the second natural language query and the contextual data, a second analytic operation query associated with the multi-dimensional dataset;
generating a second multi-dimensional data object based at least in part on the multi-dimensional dataset and executing the se

Assignees

• Arria Data2Text Ltd

Inventors

• Ponnamperuma Arachchi Kapila Anuruddha
• De Oliveira Rodrigo Gomes
• Alexander John William
• De Paiva Daniel Da Silva
• Burnett Neil Stuart

Classifications

• G06F16/243 Primary

  Natural language query formulation · CPC title