Response generation

1. A computer-implemented method, comprising:

receiving, by an electronic device operatively coupled to a processing unit, a query from a user;

obtaining, by the electronic device, a first vector of a user context of the user;

selecting, by the electronic device, one or more layers of layers of nodes from dialog trees of domains of a dialog knowledge base, wherein the nodes comprise respective dialogs;

obtaining, by the electronic device, second vectors for corresponding dialogs from corresponding nodes of the one or more layers;

generating, by the electronic device, a first set of dialog candidates based on the query, the first vector, and the second vectors;

obtaining, by the electronic device, weights of the dialogs from the dialog knowledge base;

generating, by the electronic device, a second set of dialog candidates based on the query, the first vector and the weights of the dialogs;

generating, by the electronic device, a third set of dialog candidates comprising dialog candidates present in both the first set of dialog candidates and the second set of dialog candidates;

clustering, by the electronic device, the third set of dialog candidates into clusters based on one or more hierarchal clustering techniques;

extracting, by the electronic device, respective keywords from the clusters based on a domain model;

identifying, by the electronic device, an expanded keyword that is a generalized term to represent the respective keywords based on the domain model; and

generating, by the electronic device, a response based on the third set of dialog candidates and the expanded keyword, wherein the response comprises a question to the user comprising the expanded keyword.

2. The computer-implemented method of claim 1 , wherein the user context is based on at least one of a profile, a location, a query history and a transaction history.

3. The computer-implemented method of claim 1 , wherein the query comprises feedback of the user from a last interaction of the user, and the selecting one or more layers is based on a quantity of interactions of the user.

4. The computer-implemented method of claim 1 , wherein

generating the first set of dialog candidates employs via a probabilistic algorithm.

5. The computer-implemented method of claim 1 ,

wherein the generating the first set of dialog candidates comprises determining the first set of dialog candidates based on a result of comparison between the first vector and the second vectors.

6. The computer-implemented method of claim 1 , wherein the generating the second set of dialog candidates

employs a neural network.

7. The computer-implemented method of claim 1 , further comprising:

selecting, by the electronic device, top dialogs based on the weights of the dialogs from the dialog knowledge base; and

generating, by the electronic device, the second set of dialog candidates from the top dialogs.

8. The computer-implemented method of claim 1 , further comprising:

receiving a user response to the question; and

selecting a cluster of the clusters based on the user response.

9. An electronic device, comprising:

at least one processing unit; and

a memory operatively coupled to the at least one processing unit and that stores computer executable instructions that, based on execution by the at least one processing unit, facilitate performance of operations, comprising:

receiving a query from a user;

obtaining a first vector of a user context of the user;

selecting one or more layers of layers of nodes from dialog trees of domains of a dialog knowledge base, wherein the nodes comprise respective dialogs;

obtaining second vectors for corresponding dialogs from corresponding nodes of the one or more layers;

generating a first set of dialog candidates based on the query, the first vector, and the second vectors;

obtaining weights of the dialogs from the dialog knowledge base;

generating a second set of dialog candidates based on the query, the first vector, and the weights of the dialogs;

generating a third set of dialog candidates comprising dialog candidates present in the first set of dialog candidates and the second set of dialog candidates;

clustering the third set of dialog candidates into clusters based on one or more hierarchal clustering techniques;

extracting respective keywords from the clusters based on a domain model;

identifying an expanded keyword that is a generalized term to represent the respective keywords based on the domain model; and

generating a response based on the third set of dialog candidates and the expanded keyword, wherein the response comprises a question to the user comprising the expanded keyword.

10. The electronic device of claim 9 , wherein the generating the third set of dialog candidates comprises:

generating the third set of dialog candidates further based on the user context via a search engine.

11. The electronic device of claim 9 , wherein the generating the first set of dialog candidates

employs a probabilistic algorithm.

12. The electronic device of claim 9 , wherein the generating the first set of dialog candidates comprises:

determining the first set of dialog candidates based on a result of a comparison between the first vector and the second vectors.

13. The electronic device of claim 9 , wherein the generating the second set of dialog candidates

employs a neural network.

14. The electronic device of claim 13 , wherein the generating the second set of dialog candidates comprises:

selecting top dialogs based on the weights of the dialogs; and

generating the second set of dialog candidates from the top dialogs.

15. The electronic device of claim 9 , wherein the operations further comprise:

receiving a user response to the question; and

selecting a cluster of the clusters based on the user response.

16. A computer program product for response generation, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

receive, by the processor, a query from a user;

obtain, by the processor, a first vector of a user context of the user;

select, by the processor, one or more layers of layers of nodes from dialog trees of domains of a dialog knowledge base, wherein the nodes comprise respective dialogs;

obtaining, by the processor, second vectors for corresponding dialogs from corresponding nodes of the one or more layers;

generate, by the processor, a first set of dialog candidates based on the query, the first vector, and the second vectors;

obtain, by the processor, weights of the dialogs from the dialog knowledge base;

generate, by the processor, a second set of dialog candidates based on the query, the first vector and the weights of the dialogs;

generate, by the processor, a third set of dialog candidates comprising dialog candidates present in both the first set of dialog candidates and the second set of dialog candidates;

cluster, by the processor, the third set of dialog candidates into clusters based on one or more hierarchal clustering techniques;

extract, by the processor, respective keywords from the clusters based on a domain model;

identify, by the processor, an expanded keyword that is a generalized term to represent the respective keywords based on the domain model; and

generate, by the processor, a response based on the third set of dialog candidates and the expanded keyword, wherein the response comprises a question to the user comprising the expanded keyword.

17. The computer program product of claim 16 , wherein the user context is based on at least one of a profile, a location, a query history and a transaction history.

18. The computer program product of claim 16 , wherein the program instructions are further executable by the processor to cause the processor to:

receive, by the processor, a user response to the question; and

select, by the processor, a cluster of the clusters based on the user response.

19. The computer program product of claim 18 , wherein the program instructions are further executable by the processor to cause the processor to:

generate, by the processor, an additional question to the user based on at least one dialog candidate selected from the cluster.

20. The computer program product of claim 19 , wherein the additional question comprises at least one term extracted from the at least one dialog candidate selected from the cluster.