TRANSLATING A NATURAL LANGUAGE REQUEST TO A DOMAIN SPECIFIC LANGUAGE REQUEST BASED ON MULTIPLE INTERPRETATION ALGORITHMS

• US 20190034555A1
• Filed: 07/31/2017
• Published: 01/31/2019
• Est. Priority Date: 07/31/2017
• Status: Active Application

• Current Assignee: Splunk Inc.
• Original Assignee: Splunk Inc.
• Inventors: Das, Dipock, Radon, Aungon Nag, Pochugari, Dayanand, Oliner, Adam, Padakanti, Nikesh, Srinivasabagavathar, Anand, Verma, Neeraj
• Application Number:
  US15/665,104
• Publication Number:
  US 20190034555A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/90332 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06N 5/022 Knowledge engineering; Knowledge acquisition
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 5/046 Forward inferencing; Production systems

Abstract

In various embodiments, a natural language (NL) application enables users to more effectively access various data storage systems based on NL requests. As described, the NL application includes functionality for selecting an optimal interpretation algorithm, generating a dashboard, and/or generating an alert based on an NL request. Advantageously, the operations performed by the NL application reduce the amount of time and user effort associated with accessing data storage systems and increase the likelihood of properly addressing NL requests.

Claims

1. A computer-implemented method, comprising:

executing a first interpretation algorithm to generate at least one of a first user intent and a first domain-specific language (DSL) request based on a first natural language (NL) request;

executing a second interpretation algorithm to generate at least one of a second user intent and a second DSL request based on the first NL request;

determining an optimized DSL request based on a selection criterion, the at least one of the first user intent and the first DSL request, the at least one of the second user intent and the second DSL request, and a first DSL associated with a first data storage system; and

causing the optimal DSL request to be applied to the first data storage system.

2. The computer-implemented method of claim 1, wherein executing the first interpretation algorithm generates a first confidence level, executing the second interpretation algorithm generates a second confidence level, and the selection criterion comprises a comparison between the first confidence level and the second confidence level.

3. The computer-implemented method of claim 1, wherein the selection criterion comprises a comparison between a first confidence level and a second confidence level, wherein the first confidence level indicates how effectively the at least one of the first user intent and the first DSL request conveys a meaning of the first NL request, and the second confidence level indicates how effectively the at least one of the second user intent and the second DSL request conveys the meaning of the first NL request.

4. The computer-implemented method of claim 1, wherein executing the first interpretation algorithm comprises performing at least one of an intent inference operation, a semantic similarity operation, and a disambiguation operation on the first NL request.

5. The computer-implemented method claim 1, wherein the at least one of the first user intent and the first DSL request comprises the first user intent, and determining the optimized DSL request comprises:

determining that an accuracy of the first user intent is greater than an accuracy of the at least one of the second user intent and the second DSL request based on the selection criterion; and

generating the optimal DSL request based on the first user intent and the first DSL.

6. The computer-implemented method of claim 1, wherein the first NL request includes at least one of an instruction specifying an alert and a search query.

7. The computer-implemented method of claim 1, further comprising generating a different DSL request based on a second NL request and a second DSL associated with a second data storage system, and causing the different DSL request to be applied to the second data storage system.

8. The computer-implemented method of claim 1, wherein the first data storage system stores data as a plurality of time-indexed events including respective segments of raw machine data.

9. The computer-implemented method of claim 1, wherein the first DSL comprises a pieplined search language.

10. The computer-implemented method of claim 1, further comprising:

receiving a search result associated with the optimal DSL request;

generating an NL insight based on the search result; and

causing the search result and the NL insight to be provided to a user.

11. A non-transitory computer-readable storage medium including instructions that, when executed by a processor, cause the processor to perform the steps of:

executing a first interpretation algorithm to generate at least one of a first user intent and a first domain-specific language (DSL) request based on a first natural language (NL) request;

executing a second interpretation algorithm to generate at least one of a second user intent and a second DSL request based on the first NL request;

determining an optimized DSL request based on a selection criterion, the at least one of the first user intent and the first DSL request, the at least one of the second user intent and the second DSL request, and a first DSL associated with a first data storage system; and

causing the optimal DSL request to be applied to the first data storage system.

12. The non-transitory computer-readable storage medium of claim 11, wherein executing the first interpretation algorithm generates a first confidence level, executing the second interpretation algorithm generates a second confidence level, and the selection criterion comprises a comparison between the first confidence level and the second confidence level.

13. The non-transitory computer-readable storage medium of claim 11, wherein the selection criterion comprises a comparison between a first confidence level and a second confidence level, wherein the first confidence level indicates how effectively the at least one of the first user intent and the first DSL request conveys a meaning of the first NL request, and the second confidence level indicates how effectively the at least one of the second user intent and the second DSL request conveys the meaning of the first NL request.

14. The non-transitory computer-readable storage medium of claim 11, wherein executing the first interpretation algorithm comprises performing at least one of an intent inference operation, a semantic similarity operation, and a disambiguation operation on the first NL request.

15. The non-transitory computer-readable storage medium of claim 11, wherein the at least one of the first user intent and the first DSL request comprises the first user intent, and determining the optimized DSL request comprises:

determining that an accuracy of the first user intent is greater than an accuracy of the at least one of the second user intent and the second DSL request based on the selection criterion; and

generating the optimal DSL request based on the first user intent and the first DSL.

16. The non-transitory computer-readable storage medium of claim 11, wherein the first NL request includes at least one of an instruction specifying an alert and a search query.

17. The non-transitory computer-readable storage medium of claim 11, further comprising generating a different DSL request based on a second NL request and a second DSL associated with a second data storage system, and causing the different DSL request to be applied to the second data storage system.

18. The non-transitory computer-readable storage medium of claim 11, wherein the first data storage system stores data as a plurality of time-indexed events including respective segments of raw machine data.

19. The non-transitory computer-readable storage medium of claim 11, wherein the first DSL comprises a pieplined search language.

20. The non-transitory computer-readable storage medium of claim 11, further comprising:

receiving a search result associated with the optimal DSL request;

generating an NL insight based on the search result; and

causing the search result and the NL insight to be provided to a user.

21. A computing device, comprising:

a memory that includes instructions; and

a processor that is coupled to the memory and, when executing the instructions, is configured to;

execute a first interpretation algorithm to generate at least one of a first user intent and a first domain-specific language (DSL) request based on a first natural language (NL) request;

execute a second interpretation algorithm to generate at least one of a second user intent and a second DSL request based on the first NL request;

determine an optimized DSL request based on a selection criterion, the at least one of the first user intent and the first DSL request, the at least one of the second user intent and the second DSL request, and a first DSL associated with a first data storage system; and

cause the optimal DSL request to be applied to the first data storage system.

22. The computing device of claim 21, wherein executing the first interpretation algorithm generates a first confidence level, executing the second interpretation algorithm generates a second confidence level, and the selection criterion comprises a comparison between the first confidence level and the second confidence level.

23. The computing device of claim 21, wherein the selection criterion comprises a comparison between a first confidence level and a second confidence level, wherein the first confidence level indicates how effectively the at least one of the first user intent and the first DSL request conveys a meaning of the first NL request, and the second confidence level indicates how effectively the at least one of the second user intent and the second DSL request conveys the meaning of the first NL request.

24. The computing device of claim 21, wherein executing the first interpretation algorithm comprises performing at least one of an intent inference operation, a semantic similarity operation, and a disambiguation operation on the first NL request.

25. The computing device of claim 21, wherein the at least one of the first user intent and the first DSL request comprises the first user intent, and determining the optimized DSL request comprises:

determining that an accuracy of the first user intent is greater than an accuracy of the at least one of the second user intent and the second DSL request based on the selection criterion; and

generating the optimal DSL request based on the first user intent and the first DSL.

26. The computing device of claim 21, wherein the first NL request includes at least one of an instruction specifying an alert and a search query.

27. The computing device of claim 21, wherein the processor is further configured to generate a different DSL request based on a second NL request and a second DSL associated with a second data storage system, and causing the different DSL request to be applied to the second data storage system.

28. The computing device of claim 21, wherein the first data storage system stores data as a plurality of time-indexed events including respective segments of raw machine data.

29. The computing device of claim 21, wherein the first DSL comprises a pieplined search language.

30. The computing device of claim 21, wherein the processor is further configured to:

receive a search result associated with the optimal DSL request;

generate an NL insight based on the search result; and

cause the search result and the NL insight to be provided to a user.