Virtual assistant conversations for ambiguous user input and goals

• US 9,552,350 B2
• Filed: 06/26/2014
• Issued: 01/24/2017
• Est. Priority Date: 09/22/2009
• Status: Active Grant

• Current Assignee: Verint Americas Incorporated (Verint Systems Incorporated)
• Original Assignee: Next IT Corporation (Verint Systems Incorporated)
• Inventors: Miller, Tanya M, Zartler, Mark, Brown, Fred
• Primary Examiner(s): Chau, Dung K
• Application Number:
  US14/315,852
• Publication Number:
  US 20140310005A1
• Time in Patent Office: 943 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/243 Natural language query formulation
• CPC Class Codes: G06F 16/2455 Query execution
• CPC Class Codes: G06F 16/258 Data format conversion from or to a database
• CPC Class Codes: G06F 16/3329 Natural language query formulation or dialogue systems
• CPC Class Codes: G06F 16/43 Querying
• CPC Class Codes: G06F 16/93 Document management systems
• CPC Class Codes: G06F 40/232 Orthographic correction, e.g. spell checking or vowelisation
• CPC Class Codes: G06F 40/242 Dictionaries
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06F 40/40 Processing or translation of natural language natural language analysis G06F40/20; semantic analysis G06F40/30
• CPC Class Codes: G06N 5/02 Knowledge representation; Symbolic representation
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces

Abstract

Ambiguous input of a user received during an interactive session with a virtual agent may be processed. The virtual agent may be presented via a computing device to facilitate the interactive session with the user. The user may provide the ambiguous input, which is processed to determine a response to the input. The virtual agent may provide the response to the user. The virtual agent may also carry out a goal-based dialogue where a goal to be accomplished is identified. The virtual agent may prompt the user for information related to the goal.

Claims

1. A method comprising:

causing a virtual agent to be output to facilitate an interactive session between the virtual agent and a user;

receiving initial user input during the interactive session;

processing the initial user input to determine components for the initial user input and a pattern of the components, the pattern of the components including at least one of an order of the components or a proximity of the components to each other, each of the components relating to a term in the initial user input;

determining a concept of the initial user input based at least in part on the pattern of the components;

ending the interactive session between the virtual agent and the user;

starting a subsequent interactive session between the virtual agent and the user;

receiving additional user input during the subsequent interactive session;

using the concept of the initial user input to determine a response for the additional user input; and

causing the virtual agent to output the response for the additional user input.

2. The method of claim 1, further comprising:

determining that the additional user input is ambiguous user input;

wherein the response for the additional user input is determined in response to determining that the additional user input is ambiguous user input.

3. The method of claim 2, further comprising:

determining a meaning of the additional user input by causing the virtual agent to prompt the user for further user input.

4. The method of claim 1, wherein the determining the concept of the initial user input is further based at least in part on data related to the user.

5. The method of claim 1, further comprising:

facilitating a goal-based dialogue between the virtual agent and the user to identify and accomplish a goal.

6. The method of claim 1, further comprising:

clarifying ambiguous input of the user by causing the virtual agent to prompt the user for user input.

7. The method of claim 1, wherein the additional user input comprises ambiguous user input that is associated with multiple concept-based units, each of the multiple concept-based units mapping a concept to a response.

8. The method of claim 1, further comprising:

identifying a goal associated with the initial user input; and

causing an interactive dialog to be initiated between the virtual agent and the user to acquire information for the goal.

9. One or more non-transitory computer-readable media having computer-readable instructions thereon which, when executed by one or more processors, cause the one or more processors to perform operations comprising:

enabling an interactive session between a virtual agent and a user;

receiving user input during the interactive session;

determining a concept of the user input;

starting a subsequent interactive session between the virtual agent and the user;

receiving ambiguous user input during the subsequent interactive session between the virtual agent and the user;

determining a meaning of the ambiguous user input based at least in part on the concept of the user input;

determining a response for the ambiguous user input based at least in part on the meaning of the ambiguous user input; and

causing the virtual agent to provide the response.

10. The one or more non-transitory computer-readable media of claim 9, wherein the concept is based at least in part on clarifying information that is obtained from the user.

11. The one or more non-transitory computer-readable media of claim 9, wherein the concept is based at least in part on a goal-based dialogue between the virtual agent and the user.

12. The one or more non-transitory computer-readable media of claim 9, wherein the concept is based at least in part on healthcare vocabulary.

13. The one or more non-transitory computer-readable media of claim 9, wherein the concept represents a pattern of components of the user input, the pattern of components comprising a pattern of one or more of:

a vocab term comprising at least one of a grouping of unambiguous synonyms of a term in the user input or a grouping of spelling variations of a term in the user input;

a helper term comprising a conjunction in the user input;

or

a wild card comprising a placeholder for a term in the user input.

14. The one or more non-transitory computer-readable media of claim 9, wherein the determining the meaning of the ambiguous user input includes determining that a term of the ambiguous user input has a same meaning as a term of previous user input.

15. The one or more non-transitory computer-readable media of claim 9, wherein the operations further comprise:

in response to determining the meaning of the ambiguous user input, storing the meaning of the ambiguous user input in memory; and

maintaining the meaning of the ambiguous user input in the memory during the subsequent interactive session between the virtual agent and the user.

16. The one or more non-transitory computer-readable media of claim 9, wherein the operations further comprise:

identifying a goal associated with the ambiguous user input;

assigning a goal variable to be used to complete the identified goal;

receiving a tangential request that is unrelated to the identified goal;

processing the tangential request; and

returning to the goal variable.

17. The one or more non-transitory computer-readable media of claim 9, wherein the causing the virtual assistant to provide the response comprises sending the response to a computing device for output via the computing device.

18. The one or more non-transitory computer-readable media of claim 9, wherein the causing the virtual assistant to provide the response comprises outputting the response via a computing device.

19. A system comprising:

one or more processors; and

memory communicatively coupled to the one or more processors and storing executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising;

causing a virtual agent to be output to facilitate a conversation with a user;

identifying a first goal to be accomplished, the first goal being associated with multiple variables;

receiving natural language input of the user;

determining that information for a variable of the multiple variables is missing;

causing the virtual agent to prompt the user for the information that is missing;

receiving additional natural language input;

determining that the additional natural language input relates to a second goal, distinct from the first goal;

performing a task related to the second goal; and

causing the virtual agent to return to the first goal and prompt the user again for the information that is missing for the first goal.

20. The system of claim 19, wherein the operations further comprise:

determining that the natural language input is associated with multiple meanings; and

determining a meaning of the natural language input based at least in part on a concept of previously received natural language input.

21. The system of claim 19, wherein the operations further comprise determining the information that is missing based at least in part on context of the conversation.

22. The system of claim 19, wherein the operations further comprise:

Identifying another goal to be completed to complete the first goal; and

Prompting the user for information related to the another goal.

23. A system comprising:

one or more processors; and

memory communicatively coupled to the one or more processors and storing executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising;

causing a virtual agent to be output to facilitate an interactive session between the virtual agent and a user;

receiving initial user input during the interactive session;

processing the initial user input to determine components for the initial user input and a pattern of the components, the pattern of the components including at least one of an order of the components or a proximity of the components to each other, each of the components relating to a term in the initial user input;

determining a concept of the initial user input based at least in part on the pattern of the components;

ending the interactive session between the virtual agent and the user;

starting a subsequent interactive session between the virtual agent and the user;

receiving additional user input during the subsequent interactive session;

using the concept of the initial user input to determine a response for the additional user input; and

causing the virtual agent to output the response for the additional user input.

24. The system of claim 23, wherein the operations further comprise:

determining that the additional user input is ambiguous user input;

wherein the response for the additional user input is determined in response to determining that the additional user input is ambiguous user input.

25. The system of claim 23, wherein the additional user input comprises ambiguous user input that is associated with multiple concept-based units, each of the multiple concept-based units mapping a concept to a response.

26. A system comprising:

one or more processors; and

memory communicatively coupled to the one or more processors and storing executable instructions that, when executed by the one or more processors, cause the one or more processors to perform operations comprising;

enabling an interactive session between a virtual agent and a user;

receiving user input during the interactive session;

determining a concept of the user input;

starting a subsequent interactive session between the virtual agent and the user;

receiving ambiguous user input during the subsequent interactive session between the virtual agent and the user;

determining a meaning of the ambiguous user input based at least in part on the concept of the user input;

determining a response for the ambiguous user input based at least in part on the meaning of the ambiguous user input; and

causing the virtual agent to provide the response.

27. The system of claim 26, wherein the determining the meaning of the ambiguous user input includes determining that a term of the ambiguous user input has a same meaning as a term of previous user input.

28. The system of claim 26, wherein the operations further comprise:

in response to determining the meaning of the ambiguous user input, storing the meaning of the ambiguous user input in the memory; and

maintaining the meaning of the ambiguous user input in the memory during the subsequent interactive session between the virtual agent and the user.