Analytic solution with a self-learning and context-sensitive semantic layer

Patent metadata

Field	Value
Publication number	US-9626453-B2
Application number	US-201414581046-A
Country	US
Kind code	B2
Filing date	Dec 23, 2014
Priority date	Dec 23, 2014
Publication date	Apr 18, 2017
Grant date	Apr 18, 2017

Abstract

According to a general aspect, a system includes a query engine configured to receive a query from a user via a user interface layer for obtaining data from one or more databases, determine if a keyword of the query can be mapped to at least one of a plurality of keyword mappings stored in a semantic layer, and if the keyword cannot be mapped, provide an interactive object, via the user interface layer, to learn a new keyword mapping for the keyword such that the semantics layer is updated with the new keyword mapping for future queries. The system includes a prediction engine configured to check for previous queries of the user or other users that map the query to predict a next query, and the query engine is configured to provide query results of the query and the next query as a suggestion via the user interface layer.

First claim

What is claimed is:

1. A system for dynamic data modeling that enables a self-learning and context-sensitive semantic layer to query data, the system comprising:
at least one processor;
a non-transitory computer-readable medium configured to store executable instructions that when executed by the at least one processor are configured to implement:
a query engine configured to receive a query from a user via a user interface layer for obtaining data from one or more databases, determine if a keyword of the query can be mapped to at least one of a plurality of keyword mappings stored in a semantic layer, and if the keyword cannot be mapped, provide an interactive object, via the user interface layer, to learn a new keyword mapping for the keyword such that the semantics layer is updated with the new keyword mapping for future queries; and
a prediction engine configured to check for previous queries of the user or other users that map the query to predict a next query,
the query engine configured to provide query results of the query and the next query as a suggestion via the user interface layer.

2. The system of claim 1 , wherein the query includes natural language input.

3. The system of claim 1 , wherein the query engine includes a keyword parser configured to parse the query to obtain one or more keywords.

4. The system of claim 1 , wherein the query engine includes a fuzzy logic unit configured to correct a misspelled keyword of the query based on fuzzy logic.

5. The system of claim 1 , wherein the query includes a new keyword not found in the semantic layer, the query engine including a query logic unit configured to determine that the new keyword is similar to a known keyword mapping based on thesaurus information.

6. The system of claim 1 , wherein, before the query results are provided, the query engine is configured to request feedback from the user regarding a keyword mapping of the query when a confidence level is below a threshold value.

7. The system of claim 1 , wherein the query engine is configured to provide a suggested known keyword mapping via the interactive object.

8. The system of claim 1 , wherein the query engine is configured to create a new semantic layer assignment based on input received from the interactive object.

9. The system of claim 1 , wherein the query engine is configured to receive a subsequent query from the user, and the query engine is configured to combine the query results of the query with the subsequent query to produce a combined search set to be executed on the databases.

10. A computer program product tangibly embodied on a non-transitory computer-readable storage medium and including executable code that, when executed, is configured to cause at least one processor to:
receive a query from a user for obtaining data from one or more databases;
determine if a keyword of the query can be mapped to at least one of a plurality of keyword mappings stored in a semantic layer;
if the keyword cannot be mapped, provide an interactive object, via the user interface layer, to learn a new keyword mapping for the keyword;
update the semantics layer with the new keyword mapping for future queries;
check for previous queries of the user or other users that map the query to predict a next query; and
display query results of the query and the next query to the user via the user interface layer.

11. The computer program product of claim 10 , wherein the query includes natural language input.

12. The computer program product of claim 10 , wherein the executable code that, when executed, is configured to cause the at least one processor to:
parse the query to obtain keywords including a new keyword;
enhance the keywords based on fuzzy logic; and
map the new keyword to an existing keyword mapping based on thesaurus information.

13. The computer program product of claim 10 , wherein the executable code that, when executed, is configured to cause the at least one processor to:
display the query results when a confidence level is above a threshold value, otherwise request feedback from the user regarding a keyword mapping of the query.

14. The computer program product of claim 10 , wherein the executable code that, when executed, is configured to cause the at least one processor to:
visualize the query results into at least one of a table and a chart.

15. The computer program product of claim 10 , wherein the interactive object provides a suggested known keyword mapping.

16. The computer program product of claim 10 , wherein the semantic layer also includes data relationship specifying relationships among data in the databases and data source information specifying locations of data in the databases.

17. The computer program product of claim 10 , wherein the executable code that, when executed, is configured to cause the at least one processor to:
combine the query results of the query with a subsequent query to create a combined search set.

18. A method for dynamic data modeling that enables a self-learning and context-sensitive semantic layer to query data, the method being performed by at least one processor, the method comprising:
receiving a query from a user for obtaining data from data sources, the query including a first keyword and a second keyword;
determining that the first keyword is a new keyword but is similar to a known keyword mapping stored in a semantic layer;
determining that the second keyword cannot be mapped to any known keyword mappings stored in the semantic layer;
providing an interactive object, via a user interface layer, to request a new keyword mapping for the second keyword;
updating the semantics layer with the new keyword mapping for future queries; and
obtaining query results corresponding to the query having the first keyword and the second keyword; and
displaying the query results via the user interface layer.

19. The method of claim 18 , wherein the query includes natural language input.

20. The method of claim 18 , further comprising:
receiving a subsequent query;
combining the query results of the query with the subsequent query to obtain a combined search; and
executing the combined search on the data sources.

Assignees

• Koerner Dennis
• Linke Andreas
• Sap Se

Inventors

• Koerner Dennis
• Linke Andreas

Classifications

• G06F16/90324 Primary

  using system suggestions · CPC title

• G06F16/243

  Natural language query formulation · CPC title

• G06F17/30401

  Physics · mapped topic

• G06F17/3097 Primary

  Physics · mapped topic