Interaction driven artificial intelligence system and uses for same, including travel or real estate related contexts
Artificial intelligence systems utilizing a chat driven interface that includes a chat portion and a display portion are disclosed. The user may interact with the chat portion of the interface in a human understandable and comfortable manner (e.g., through natural language). The artificial intelligence system uses the context of the interactions in the chat portion of the interface to drive the presentation of content in the display portion interface and the chat portion such that the content presented in the display portion of the interface reflects the intents or concepts expressed in the chat portion of the interface, and the chat portion and the display portion of the interface are kept in synchronicity.
1. A chat driven artificial intelligence system, comprising:
a processor; and
a non-transitory computer-readable medium comprising instructions for:
initializing a chat system widget executing on a user device;
receiving a request from the chat system widget at the user device, wherein the request includes a current interaction entered by a user in a chat portion of a chat driven interface displayed at the user device, and wherein the request represents at least a portion of a current session with the user;
determining an intent or an entity from the request from the chat system widget;
retrieving one or more historical session(s) associated with the user of the user device, wherein the one or more historical session(s) include data related to one or more previous requests from the user;
determining an intent or an entity from the historical session(s);
formulating a response to the request that includes content related to the entity or the intent from the historical session(s) and content related to the entity or the determined intent from the current session; and
sending the formulated response to the chat system widget to be displayed on a user device, wherein the response includes a return frame and wherein the response is associated with the current session and the historical session(s) associated with the user.
2. The system of claim 1 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
creating a working frame from the one or more historical session(s) that includes the at least one intent or the at least one entity.
3. The system of claim 2 ,
wherein when determining an intent from the request from the chat system widget, the request is sent to a natural language processing extraction service;
wherein the response from the natural language processing extraction service includes a determined intent and a determined concept;
and
wherein evaluating the determined concept from the current session and one or more concepts from the historical session(s) to determine a canonical concept for the working frame.
4. The system of claim 3 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
receiving, by a data service, the canonical concept for the working frame;
and
sending, by the data service, any one or more of: content related to the canonical concept for the working frame and a set of follow-up rules to the chat system widget.
5. The system of claim 1 ,
wherein a natural language processing extraction module determines the intent or the entity from the request from the chat system widget; and
wherein the natural language processing extraction module includes a pre-processing module that receives the request and the current session and creates a working frame.
6. The system of claim 1 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
determining, by a follow-up module, if the request received from the chat system widget at the user device is a complete request; and
wherein the follow-up module includes a plurality of follow up rules for determining if a follow up question is needed to obtain more data from the user.
7. The system of claim 6 ,
wherein the follow up module evaluates the entity or the intent from the historical session(s) and the entity or the determined intent from the current session to determine if more information is needed from the user to complete the request.
8. The system of claim 1 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
sending to a data service the entity or the intent from the historical session(s) and the entity or the determined intent from the current session to determine content in response to the request; and
wherein the data service is queried based on a geographical region extracted from the current session or one of the one or more historical session(s) with the chat portion of the chat interface.
9. The system of claim 1 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
storing the current session as a historical session for the user.
10. The system of claim 1 ,
wherein the formulated response is related to travel; and
wherein the non-transitory computer-readable medium further comprises instructions for:
updating a map presented in a map display area of the chat driven interface to include at least a portion of the formulated response that is related to travel.
11. A method for a chat driven artificial intelligence system, the method comprising:
initializing a chat system widget executing on a user device;
receiving a request from the chat system widget at the user device, wherein the request includes a current interaction entered by a user in a chat portion of a chat driven interface displayed at the user device, and wherein the request represents at least a portion of a current session with the user;
determining an intent or an entity from the request from the chat system widget;
retrieving one or more historical session(s) associated with the user of the user device, wherein the one or more historical session(s) include data related to one or more previous requests from the user;
determining an intent or an entity from the historical session(s);
formulating a response to the request that includes content related to the entity or the intent from the historical session(s) and content related to the entity or the determined intent from the current session; and
sending the formulated response to the chat system widget to be displayed on a user device, wherein the response includes a return frame and wherein the response is associated with the current session and the historical session(s) associated with the user.
12. The method of claim 11 , further comprising:
creating a working frame from the one or more historical session(s) that includes the at least one intent or the at least one entity.
13. The method of claim 12 ,
wherein when determining an intent from the request from the chat system widget, the request is sent to a natural language processing extraction service;
wherein the response from the natural language processing extraction service includes a determined intent and a determined concept;
and
wherein evaluating the determined concept from the current session and one or more concepts from the historical session(s) to determine a canonical concept for the working frame.
14. The method of claim 13 , further comprising
receiving, by a data service, the canonical concept for the working frame; and
sending, by the data service, any one or more of: content related to the canonical concept for the working frame and a set of follow-up rules to the chat system widget.
15. The method of claim 11 ,
wherein a natural language processing extraction module determines the intent or the entity from the request from the chat system widget; and
wherein the natural language processing extraction module includes a pre-processing module that receives the request and the current session and creates a working frame.
16. The method of claim 11 , further comprising:
determining, by a follow-up module, if the request received from the chat system widget at the user device is a complete request; and
wherein the follow-up module includes a plurality of follow up rules for determining if a follow up question is needed to obtain more data from the user.
17. The method of claim 16 ,
wherein the follow up module evaluates the entity or the intent from the historical session(s) and the entity or the determined intent from the current session to determine if more information is needed from the user to complete the request.
18. The method of claim 11 , further comprising:
sending to a data service the entity or the intent from the historical session(s) and the entity or the determined intent from the current session to determine content in response to the request; and
wherein the data service is queried based on a geographical region extracted from the current session or one of the one or more historical session(s) with the chat portion of the chat interface.
19. The method of claim 11 ,
wherein the non-transitory computer-readable medium further comprises instructions for:
storing the current session as a historical session for the user.
20. The method of claim 11 ,
wherein the formulated response is related to travel; and
wherein the non-transitory computer-readable medium further comprises instructions for:
updating a map presented in a map display area of the chat driven interface to include at least a portion of the formulated response that is related to travel.