Claims:We claim:

1. A method of determining response to queries in virtual assistance system, the method comprising:
receiving, by a virtual assistance system (101), data comprising a plurality of interaction between one or more users and one or more real agents for resolving a query, wherein the data is classified into user data and real agent data;
classifying, by the virtual assistance system (101), one or more entities and intent identified from each of the classified user data and the real agent data using a predefined domain model, wherein the one or more entities and the intent are combined to identify a plurality of a sequence of resolution data;
determining, by the virtual assistance system (101), a first set of resolution data indicative of mandatory information required for resolving the query and a second set of resolution data indicative of supplementary information for resolving the query, based on the classification of the one or more entities and the intent; and
clustering, by the virtual assistance system (101), each of the plurality of sequence of resolution data into one or more category associated with type of users, based on at least one of the first set of resolution data and the second set of resolution data, one or more parameters associated with the one or more users and parameters associated with historical resolution data used for responding to the query.

2. The method as claimed in claim 1, wherein the predefined domain model is trained using a plurality of training data associated with user and agent interaction in one or more domains by machine learning techniques, wherein the plurality of training data is annotated with one or more entities and intent.

3. The method as claimed in claim 1, wherein each of the first set of resolution data and the second set of resolution data are classified based on number of times each data appeared during interaction with different users.

4. The method as claimed in claim 1, wherein the one or more category associated with type of users depends upon knowledge of the users in a domain related to the query.

5. The method as claimed in claim 1, wherein the one or more parameters associated with the one or more users comprises location of users, skills, domain background, age, domain proficiency and time-taken by each user to follow and complete resolution provided by the one or more real agents.

6. The method as claimed in claim 1, wherein the parameters associated with the associated with historical resolution data comprises time taken for resolving for each user query interaction, number of total steps provided by each real agent during the interaction and success rate of the resolution.

7. The method as claimed in claim 1 further comprising storing each of the plurality of sequence of resolution data based on the one or more category in a database associated with the virtual assistance system.

8. A virtual assistance system (101) for determining response to queries, comprising:
a processor (115); and
a memory (111) communicatively coupled to the processor (115), wherein the memory (111) stores processor instructions, which, on execution, causes the processor (115) to:
receive data comprising a plurality of interaction between one or more users and one or more real agents for resolving a query, wherein the data is classified into user data and real agent data;
classify one or more entities and intent identified from each of the classified user data and the real agent data using a predefined domain model, wherein the one or more entities and the intent are combined to identify a plurality of a sequence of resolution data;
determine a first set of resolution data indicative of mandatory information required for resolving the query and a second set of resolution data indicative of supplementary information for resolving the query, based on the classification of the one or more entities and the intent; and
cluster each of the plurality of sequence of resolution data into one or more category associated with type of users, based on at least one of the first set of resolution data and the second set of resolution data, one or more parameters associated with the one or more users and parameters associated with historical resolution data used for responding to the query.
9. The virtual assistance system (101) as claimed in claim 8, wherein the processor trains the predefined domain model using a plurality of training data associated with user and agent interaction in one or more domains by machine learning techniques, wherein the plurality of training data is annotated with one or more entities and intent.
10. The virtual assistance system (101) as claimed in claim 8, wherein the processor classifies each of the first set of resolution data and the second set of resolution data based on number of times each data appeared during interaction with different users.

Dated this 10th day of July, 2019

Madhusudan S T
Of K&S Partners
Agent for the Applicant
IN/PA-1297
, Description:TECHNICAL FIELD
The present subject matter is related in general to adaptive virtual assistance systems and learning, more particularly, but not exclusively to a method and system for determining response to queries in virtual assistance system.