Cognitive switching logic for multiple knowledge domains

1. A method for providing a predicted response to a user, the method comprising:

training a neural network with an initial set of natural language information for a set of multiple knowledge domains of a domains model;

retraining the neural network based on user feedback and user persona metadata to obtain a trained neural network for the set of multiple knowledge domains;

receiving a message from a user, the message having unstructured natural language information;

processing, using cognitive switching logic, the unstructured natural language information of the message and information from previous messages from the same user, wherein the cognitive switching logic is connected to the set of multiple knowledge domains of the domains model, and comprises the trained neural network;

identifying, using the cognitive switching logic, a context of the unstructured natural language information of the message based on the information from the previous messages;

identifying, using the cognitive switching logic, at least one knowledge domain of the multiple knowledge domains which contains a response to the message, based on the identified context of the unstructured natural language information of the message and on user persona information, wherein the cognitive switching logic includes the trained neural network to identify the at least one knowledge domain of the set of multiple knowledge domains of the domains model which contains a response to the message;

retrieving a respective response from each identified knowledge domain;

obtaining, by the cognitive switching logic, feedback, based on retrieving more than one respective response, to refine identifying the at least one knowledge domain until only one response is retrieved; and

presenting the one response to the user.

2. The method of claim 1 , further comprising:

receiving user feedback from the user based on the presented response; and

refining the cognitive switching logic based on the user feedback.

3. The method of claim 2 , wherein refining the cognitive switching logic based on the user feedback includes modifying the identification of the context of the unstructured natural language information.

4. The method of claim 2 , wherein:

in response to receiving user feedback indicating that the presented response is incorrect, the cognitive switching logic generates a penalizing factor to be utilized by the neural network.

5. A computerized system for providing a predicted response to a user, the computerized system including a processor configured to perform a method, the method comprising:

training a neural network with an initial set of natural language information for a set of multiple knowledge domains of a domains model;

retraining the neural network based on user feedback and user persona metadata to obtain a trained neural network for the set of multiple knowledge domains;

receiving a message from a user, the message having unstructured natural language information;

processing, using cognitive switching logic, the unstructured natural language information of the message and information from previous messages from the same user, wherein the cognitive switching logic is connected to the set of multiple knowledge domains of the domains model, and comprises the trained neural network;

identifying, using the cognitive switching logic, a context of the unstructured natural language information of the message based on the information from the previous messages;

identifying, using the cognitive switching logic, at least one knowledge domain of the multiple knowledge domains which contains a response to the message, based on the identified context of the unstructured natural language information of the message and on user persona information, wherein the cognitive switching logic includes the trained neural network to identify the at least one knowledge domain of the set of multiple knowledge domains of the domains model which contains a response to the message;

retrieving a respective response from each identified knowledge domain;

obtaining, by the cognitive switching logic, feedback, based on retrieving more than one respective response, to refine identifying the at least one knowledge domain until only one response is retrieved; and

presenting the one response to the user.

6. The computerized system of claim 5 , the method further comprising:

receiving user feedback from the user based on the presented response; and

refining the cognitive switching logic based on the user feedback.

7. The computerized system of claim 6 , wherein refining the cognitive switching logic based on the user feedback includes modifying the identification of the context of the unstructured natural language information.

8. The computerized system of claim 7 , wherein:

in response to receiving user feedback indicating that the presented response is incorrect, the cognitive switching logic generates a penalizing factor to be utilized by the neural network.

9. A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions being executable by a computer system to cause the computer system to perform a method for providing a predicted response to a user, the method comprising:

training a neural network with an initial set of natural language information for a set of multiple knowledge domains of a domains model;

retraining the neural network based on user feedback and user persona metadata to obtain a trained neural network for the set of multiple knowledge domains;

receiving a message from a user, the message having unstructured natural language information;

processing, using cognitive switching logic, the unstructured natural language information of the message and information from previous messages from the same user, wherein the cognitive switching logic is connected to the set of multiple knowledge domains of the domains model, and comprises the trained neural network;

identifying, using the cognitive switching logic, a context of the unstructured natural language information of the message based on the information from the previous messages;

identifying, using the cognitive switching logic, at least one knowledge domain of the multiple knowledge domains which contains a response to the message, based on the identified context of the unstructured natural language information of the message and on user persona information, wherein the cognitive switching logic includes the trained neural network to identify the at least one knowledge domain of the set of multiple knowledge domains of the domains model which contains a response to the message;

retrieving a respective response from each identified knowledge domain;

obtaining, by the cognitive switching logic, feedback, based on retrieving more than one respective response, to refine identifying the at least one knowledge domain until only one response is retrieved; and

presenting the one response to the user.

10. The computer program product of claim 9 , the method further comprising:

receiving user feedback from the user based on the presented response; and

refining the cognitive switching logic based on the user feedback.

11. The computer program product of claim 10 , wherein refining the cognitive switching logic based on the user feedback includes modifying the identification of the context of the unstructured natural language information.

12. The computer program product of claim 11 , wherein:

in response to receiving user feedback indicating that the presented response is incorrect, the cognitive switching logic generates a penalizing factor to be utilized by the neural network.