Query response using mapping to parameterized report

Patent metadata

Field	Value
Publication number	US-10318528-B2
Application number	US-201615054012-A
Country	US
Kind code	B2
Filing date	Feb 25, 2016
Priority date	Feb 25, 2016
Publication date	Jun 11, 2019
Grant date	Jun 11, 2019

Abstract

A mechanism for responding to a natural language query. Linguistic parsing is applied to the natural language query to generate a set of language element(s). One or more of those language elements are mapped to parameters and/or data elements of a parameterized report, which is selected for use in responding to the query based on the mapping. The selection of the parameterized report and/or mapping to the parameterized report may use the schema of the underlying dataset that the parameterized report uses in order to perform more precise mapping and selection. The selected parameterized report is then used to construct structured data that is responsive to the query. For instance, the structured data could include visualizations. In this manner, already existing parameterized reports that were previously used to generated structured data in the form of reports, may again be used in order to respond to natural language queries.

First claim

What is claimed is:

1. A system comprising:
a dataset having a plurality of data elements;
a plurality of parameterized reports that are constructed from structured data based on at least the dataset, the structured data for each of the plurality of parameterized reports having a corresponding set of data elements;
a natural language processing engine configured to apply a linguistic parsing to natural language queries to generate corresponding language elements;
a mapping component that, for each respective report of at least some of the plurality of parameterized reports:
maps some of the language elements to at least some of the data elements of the respective parameterized report; and
assigns a mapping score to the respective parameterized report;

a report selection component that, for each of at least some natural language queries and based on the mapping scores of the mapping component, selects one or more of the plurality of parameterized reports to respond to the natural language query;
a report modification component that, when it is determined the selected one or more of the plurality of parameterized reports is insufficient to fully respond to the natural language query, generates a derived report based on at least one of the selected parameterized reports; and
a structuring component that uses at least the selected one or more of the plurality of parameterized reports, and any derived reports necessary to respond to the natural language query, to generate structured data to respond to the natural language query.

2. The system in accordance with claim 1 , the report selection component also using a schema of the particular dataset to select the one or more parameterized reports to use to respond to the natural language query for at least some of the natural language queries.

3. The system in accordance with claim 1 , the mapping component also using a schema of the particular data set to perform the mapping for at least some of the natural language queries.

4. The system in accordance with claim 1 , the mapping component also using one or more relationships between language elements to perform the mapping for at least some of the natural language queries.

5. The system in accordance with claim 1 , the derived report having one or more parameters and/or data elements removed from the corresponding selected parameterized report.

6. The system in accordance with claim 5 , the one or more parameters and/or data elements being removed including one or more particular parameters being removed, the derived report having one or more data elements added to the corresponding selected parameterized report to replace the one or more particular parameters that had been removed from the corresponding selected parameterized report.

7. The system in accordance with claim 1 , the derived report having one or more parameters and/or data elements added to the corresponding selected parameterized report.

8. The system in accordance with claim 1 , the derived report replacing one or more parameters and/or data elements of the corresponding selected parameterized report.

9. The system in accordance with claim 1 , the corresponding selected parameterized report being a visualized report, the derived report also being a visualized report that has a different visualization than the corresponding selected parameterized report.

10. The system in accordance with claim 1 , at least some of the plurality of parameterized reports being visualizations.

11. The system in accordance with claim 1 , the mapping performed by the mapping component including mapping to parameters based on described parameters of at least some of the plurality of parameterized reports.

12. A method for responding to a natural language query, the method comprising:
obtaining a dataset having a plurality of data elements;
obtaining a plurality of parameterized reports that are constructed from structured data based on at least the dataset, the structured data for each of the plurality of parameterized reports having a corresponding set of data elements;
accessing a natural language query;
apply computerized linguistic parsing to the natural language query to generate a set of natural language elements;
for each of at least some of the plurality of parameterized reports:
mapping some of the natural language elements to at least some of the data elements of the parameterized report; and
assigning a mapping score to the parameterized report;

based on the assigned mapping scores, selecting one of the plurality of parameterized reports for use in responding to the natural language query;
determining that the selected one of the plurality of parameterized reports is insufficient to fully respond to the natural language query;
generating a derived report based on at least the selected one of the plurality of parameterized reports; and
using the selected one of the plurality of parameterized reports, and any derived reports necessary to respond to the natural language query, to generated structured data that responds to the natural language query.

13. The method in accordance with claim 12 , further comprising using a schema of the dataset, in addition to the assigned mapping score, to select the parameterized report to use to respond to the natural language query.

14. The method in accordance with claim 12 , the mapping further including using a schema of the particular dataset to perform the mapping.

15. The method in accordance with claim 12 , the mapping further including using one or more relationships between language elements to perform the mapping.

16. The method in accordance with claim 1 , the selected parameterized report being a visualization, the derived parameterized reporting being a different visualization.

17. The method in accordance with claim 12 , the selected parameterized report being a visualization.

18. A computer program product comprising one or more computer-readable storage media having thereon computer-executable instructions that are structured such that, when executed by one or more processors of a computing system, the computing system is configured to instantiate and/or operate the following:
a natural language processing engine configured to apply a linguistic parsing to natural language queries to generate corresponding sets of language elements;
a mapping component that performs the following for each respective report of a plurality of parameterized reports, wherein each parameterized report includes a set of data elements corresponding to a set of data elements of a dataset:
maps some of the generated language elements to at least some of the data elements of the respective parameterized report; and
assigns a mapping score to the respective parameterized report;

a report selection component that, based on the assigned mapping scores, selects a parameterized report to use to respond to the natural language query;
a report modification component that, when it is determined the selected parameterized report is insufficient to fully respond to the natural language query, generates a derived report based at least on the selected parameterized report; and
a structuring component that uses at least the selected parameterized report, and the derived report, to generate structured data to respond to the natural language query.

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Hays Christopher A
• Baumgartner Patrick J
• Meyers Aaron
• Wu Jingxiao

Classifications

• G06F16/243 Primary

  Natural language query formulation · CPC title

• G06F16/248

  Presentation of query results · CPC title

• G06F16/2272

  Management thereof · CPC title

• G06F16/24522 Primary

  Translation of natural language queries to structured queries · CPC title