Method of and system for using conversation state information in a conversational interaction system
View Patent ↗A method of using conversation state information in a conversational interaction system is disclosed. A method of inferring a change of a conversation session during continuous user interaction with an interactive content providing system includes receiving input from the user including linguistic elements intended by the user to identify an item, associating a linguistic element of the input with a first conversation session, and providing a response based on the input. The method also includes receiving additional input from the user and inferring whether or not the additional input from the user is related to the linguistic element associated with the conversation session. If related, the method provides a response based on the additional input and the linguistic element associated with the first conversation session. Otherwise, the method provides a response based on the second input without regard for the linguistic element associated with the first conversation session.
1. A computer-implemented method of inferring a change of a conversation session during continuous user interaction with an interactive content providing system having one or more processors, the method comprising:
providing access to a set of content items, each of the content items having associated metadata stored in an electronically readable medium that describes the corresponding content item;
receiving at one or more of the processors a first input from a user, the first input including linguistic elements intended by the user to identify at least one desired content item;
associating by one or more of the processors at least one linguistic element of the first input with a first conversation session;
providing by one or more of the processors a first response based on the first input and based on the metadata associated with the content items;
receiving at one or more of the processors a second input from the user;
inferring by one or more of the processors whether or not the second input from the user is related to the at least one linguistic element associated with the first conversation session;
upon a condition in which the second input is inferred to relate to the at least one linguistic element associated with the first conversation session, providing by one or more of the processors a second response based on the metadata associated with the content items, the second input, and the at least one linguistic element of the first input associated with the first conversation session; and
upon a condition in which the second input is inferred to not relate to the at least one linguistic element associated with the first conversation session, providing by one or more of the processors a second response based on the metadata associated with the content items and the second input.
2. The method of claim 1 , wherein the inferring by the one or more processors whether or not the second input from the user is related to the at least one linguistic element associated with the first conversation session is based on a determination of whether or not the second input contains a linguistic linking element.
3. The method of claim 2 , wherein the linguistic linking element is at least one of a pronoun and a syntactic expletive.
4. The method of claim 2 , wherein the linguistic linking element is at least one of a noun and a named entity.
5. The method of claim 2 , wherein the inferring by the one or more processors concludes the second input from the user is related upon a condition in which the second input contains a linguistic linking element.
6. The method of claim 5 , wherein the linguistic linking element is at least one of a pronoun and a syntactic expletive.
7. The method of claim 5 , wherein the linguistic linking element is at least one of a noun and a named entity.
8. The method of claim 2 , wherein the inferring by the one or more processors concludes the second input from the user is not related upon a condition in which the second input does not contain a linguistic linking element.
9. The method of claim 8 , wherein the linguistic linking element is at least one of a pronoun and a syntactic expletive.
10. The method of claim 8 , wherein the linguistic linking element is at least one of a noun and a named entity.
11. The method of claim 2 , further comprising, upon a condition in which the second input does not contain a linguistic linking element, the following:
determining by the one or more processors a measure of relatedness between (i) linguistic elements of the second input and (ii) the at least one linguistic element associated with the first conversation session based on the metadata associated with the content items;
upon a condition in which the measure of relatedness is equal to or greater than a threshold value, the inferring by the one or more processors concludes the second input is related to the at least one linguistic element associated with the first conversation session; and
upon a condition in which the measure of relatedness is less than the threshold value, the inferring by the one or more processors concludes the second input is not related to the at least one linguistic element associated with the first conversation session.
12. The method of claim 11 , the metadata associated with the content items including a mapping of relationships between entities associated with the content items, and the determining by the one or more processors the measure of relatedness including analyzing by the one or more processors the mapping of relationships.
13. The method of claim 1 , wherein the inferring by the one or more processors whether or not the second input from the user is related to the at least one linguistic element associated with the first conversation session includes:
identifying by the one or more processors a linguistic element associated with the first conversation session that identifies at least one entity;
identifying by the one or more processors a linguistic linking element of the second input;
determining by the one or more processors whether or not the linguistic linking element of the second input is a suitable link to the linguistic element associated with the first conversation session that identifies at least one entity;
upon a condition in which the linguistic linking element of the second input is a suitable link, concluding by the one or more processors the second input from the user is related to the at least one linguistic element associated with the first conversation session; and
upon a condition in which the linguistic linking element of the second input is not a suitable link, concluding by the one or more processors the second input from the user is not related to the at least one linguistic element associated with the first conversation session.
14. The method of claim 13 , wherein the linguistic element associated with the first conversation session that identifies at least one entity is at least one of a noun and a gerund.
15. The method of claim 14 , wherein the linguistic linking element of the second input is at least one of a pronoun and a syntactic expletive.
16. The method of claim 14 , wherein the linguistic linking element of the second input is at least one of a noun and a named entity.
17. The method of claim 1 , wherein the inferring by the one or more processors whether or not the second input from the user is related to the at least one linguistic element associated with the first conversation session includes:
determining by the one or more processors whether or not the second input includes a linguistic element that identifies at least one entity;
upon a condition in which the second input does not include a linguistic element that identifies at least one entity, determining by the one or more processors whether or not the second input is an appropriate response to the first response;
upon a condition in which the second input is an appropriate response to the first response, concluding by the one or more processors the second input from the user is related to the at least one linguistic element associated with the first conversation session; and
upon a condition in which the second input is not an appropriate response to the first response, concluding by the one or more processors the second input from the user is not related to the at least one linguistic element associated with the first conversation session.
18. The method of claim 17 , wherein the linguistic element included in the second input that identifies at least one entity is at least one of a noun and a gerund.
19. The method of claim 1 , upon the condition in which the second input is inferred to relate to the at least one linguistic element associated with the first conversation session, the providing by the one or more processors the second response including substituting by the one or more processors the at least one linguistic element associated with the first conversation session in place of at least one linguistic element of the second input.
20. The method of claim 1 , further comprising:
associating by the one or more processors at least one linguistic element of the first response with the first conversation session; and
upon a condition in which the second input is inferred to relate to the at least one linguistic element associated with the first conversation session, further basing the second response on the at least one linguistic element of the first response associated with the first conversation session.
21. The method of claim 20 , upon the condition in which the second input is inferred to relate to the at least one linguistic element associated with the first conversation session, the providing by the one or more processors the second response including substituting the at least one linguistic element associated with the first conversation session in place of at least one linguistic element of the second input.
22. The method of claim 1 , the metadata associated with the content items including a mapping of relationships between entities associated with the content items.
23. The method of claim 1 , further comprising:
determining by the one or more processors that a portion of at least one of the first input from the user and the second input from the user contains an ambiguous identifier, the ambiguous identifier intended by the user to identify, at least in part, the at least one desired content item; and
inferring by the one or more processors a meaning for the ambiguous identifier based on matching portions of the at least one of the first input from the user and the second input from the user to preferences of the user described by a user preference signature;
the providing by the one or more processors a second response including selecting content items from the set of content items based on comparing the inferred meaning of the ambiguous identifier with metadata associated with the content items.