Disambiguating user intent in conversational interaction system for large corpus information retrieval

1. A method comprising:

receiving a first input from a user, the first input intended to identify a first set of entities; presenting the first set of entities based on the first input;

receiving a second input from the user in response to presenting the first set of entities, the second input intended to identify a second set of entities in an information domain space;

determining, based on comparing the second input to information in the information domain space, an ambiguity level of the second input

determining whether the ambiguity level of the second input exceeds a threshold value; and

in response to determining that the ambiguity level of the second input exceeds the threshold value:

receiving metadata describing each entity of the second set of entities;

receiving structural knowledge, the structural knowledge showing semantic links between each entity of the first set of entities and the second set of entities;

determining a disambiguated input based on the metadata of the second set of entities and the structural knowledge; and

selecting the second set of entities based on the disambiguated input.

2. The method of claim 1 , wherein the determined disambiguated input indicates intent of the user, wherein the disambiguated input is stored in a database.

3. The method of claim 1 , further comprising:

ordering the second set of entities based on the structural knowledge;

constructing a natural language message stating that an entity of the ordered second set of entities matches the disambiguated input; and

generating for display a notification for the user comprising the natural language message.

4. The method of claim 1 , further comprising:

in response to determining that at least the first set of entities and the structural knowledge do not disambiguate the second input:

presenting a disambiguating question to the user;

receiving a third input from the user in response to the disambiguating question; and

determining a disambiguated input further based on the third input.

5. The method of claim 1 , further comprising:

in response to determining that the first set of entities and the structural knowledge do not disambiguate the second input:

generating for display a notification for the user that multiple options exist for the second input.

6. The method of claim 1 , wherein the disambiguated input is further determined based on a user preference, a location of the user, and a time of the second input.

7. The method of claim 6 , wherein determining the disambiguated input further comprises:

determining whether the user preference is available; and

in response to determining that the user preference is available, determining the disambiguated input based on the user preference.

8. The method of claim 1 , wherein determining the disambiguated input is further based on user history.

9. A method comprising:

receiving a first input from a user, the first input intended to identify a first set of entities; presenting the first set of entities based on the first input

receiving a second input from the user in response to presenting the first set of entities, the second input intended to identify a second set of entities;

determining whether an ambiguity level of the second input exceeds a threshold value; and

in response to determining that the ambiguity level of the second input exceeds the threshold value:

receiving metadata describing each entity of the second set of entities;

receiving structural knowledge, the structural knowledge showing semantic links between each entity of the first set of entities and the second set of entities;

determining a disambiguated input based on the metadata of the second set of entities and the structural knowledge;

selecting the second set of entities based on the disambiguated input

receiving a third input from the user, the third input intended to identify a third set of entities;

receiving metadata describing the particular entity;

determining whether an ambiguity level of the third input exceeds another threshold value; and

in response to determining that the ambiguity level of the third input exceeds the another threshold value:

determining a disambiguated input based on at least one of the structural knowledge or the first set of entities; and

selecting the third set of entities based on comparing the disambiguated input with the metadata describing the third input.

10. The method of claim 1 , wherein selecting the second set of entities comprises:

comparing the disambiguated input with the metadata describing each respective entity of the second set of entities;

determining, based on the comparing, that no entities in the second set of entities satisfy the disambiguated input; and

selecting a subset of entities that satisfies a portion of the disambiguated input.

11. A system comprising:

control circuitry configured to:

receive a first input from a user, the first input intended to identify a first set of entities;

present the first set of entities based on the first input;

receive a second input from the user in response to presenting the first set of entities, the second input intended to identify a second set of entities in an information domain space;

determine, based on comparing the second input to information in the information domain space, an ambiguity level of the second input

determine whether the ambiguity level of the second input exceeds a threshold value; and

in response to determining that the ambiguity level of the second input exceeds the threshold value:

receive metadata describing each entity of the second set of entities;

receive structural knowledge, the structural knowledge showing semantic links between each entity of the first set of entities and the second set of entities;

determine a disambiguated input based on the metadata of the second set of entities and the structural knowledge; and

select the second set of entities based on the disambiguated input; and

input/output circuitry configured to:

present the selected set of entities for display.

12. The system of claim 11 , wherein the determined disambiguated input indicates intent of the user, wherein the disambiguated input is stored in a database.

13. The system of claim 11 , wherein the control circuity is further configured to:

order the second set of entities based on the structural knowledge;

construct a natural language message stating that an entity of the ordered second set of entities matches the disambiguated input; and

generate for display a notification for the user comprising the natural language message.

14. The system of claim 11 , wherein the control circuity is further configured to:

in response to determining that at least the first set of entities and the structural knowledge do not disambiguate the second input:

present a disambiguating question to the user;

receive a third input from the user in response to the disambiguating question; and

determine a disambiguated input further based on the third input.

15. The system of claim 11 , wherein the control circuity is further configured to:

in response to determining that the first set of entities and the structural knowledge do not disambiguate the second input:

generate for display a notification for the user that multiple options exist for the second input.

16. The system of claim 11 , wherein the disambiguated input is further determined based on a user preference, a location of the user, and a time of the second input.

17. The system of claim 16 , wherein the control circuitry, when determining the disambiguated input, is further configured to:

determine whether the user preference is available; and

in response to determining that the user preference is available, determine the disambiguated input based on the user preference.

18. The system of claim 11 , wherein determining the disambiguated input is further based on user history.

19. The system of claim 11 , wherein the control circuity is further configured to:

receive a third input from the user, the third input intended to identify a third set of entities;

receive metadata describing the particular entity;

determine whether an ambiguity level of the third input exceeds another threshold value; and

in response to determining that the ambiguity level of the third input exceeds the another threshold value:

determine a disambiguated input based on at least one of the structural knowledge or the first set of entities; and

select the third set of entities based on comparing the disambiguated input with the metadata describing the third input.

20. The system of claim 11 , wherein the control circuity, when selecting the second set of entities, is further configured to:

compare the disambiguated input with the metadata describing each respective entity of the second set of entities;

determine, based on the comparing, that no entities in the second set of entities satisfy the disambiguated input; and

select a subset of entities that satisfies a portion of the disambiguated input.