US 11,966,705 B2
Systems and methods for bot dialog delegation
Allen Detmer, Boston, MA (US); and Maria Lucena, Boston, MA (US)
Assigned to FMR LLC, Boston, MA (US)
Filed by FMR LLC, Boston, MA (US)
Filed on Nov. 13, 2020, as Appl. No. 17/097,174.
Claims priority of provisional application 62/934,897, filed on Nov. 13, 2019.
Prior Publication US 2021/0142009 A1, May 13, 2021
Int. Cl. G06F 40/35 (2020.01); G06N 5/04 (2023.01); H04L 51/02 (2022.01)
CPC G06F 40/35 (2020.01) [G06N 5/04 (2013.01); H04L 51/02 (2013.01)] 10 Claims
OG exemplary drawing
 
1. A computerized method for switching between natural language processing units during an interaction using a dialog delegation processor, the method comprising:
receiving, by a server computing device, a plurality of intent definitions, wherein each intent definition is associated with a different one of a plurality of natural language processors connected to the server computing device and comprises a set of dialog responses to an intent associated with a user utterance;
receiving, by the server computing device, a dialog utterance from the user via a user device, the dialog utterance associated with a first dialog flow and a conversation state;
invoking, by the server computing device, each of the plurality of natural language processors using the dialog utterance until a first one of the natural language processors classifies the dialog utterance with a general intent of the user, including incrementing an incomprehension counter for each natural language processor that does not classify the dialog utterance with a general intent of the user and requesting a rephrased dialog utterance from the user when the incomprehension count exceeds a predetermined threshold;
comparing, by the server computing device, the general intent of the user with the plurality of intent definitions to identify a dialog response to the general intent, the identified dialog response comprising a delegation to a second dialog flow that references a second one of the plurality of natural language processors and wherein the second natural language processor is referenced in the delegation according to a first specific intent corresponding to the general intent of the user;
invoking, by the server computing device, the second natural language processor based upon the identified dialog response to execute the second dialog flow;
receiving, by the second natural language processor, the conversation state and the dialog utterance;
generating, by the second natural language processor, a dialog response to the dialog utterance;
updating, by the second natural language processor, the conversation state;
returning, by the second natural language processor, the dialog response and updated conversation state to the first natural language processor via the server computing device;
returning, by the first natural language processor, a response to the server computing device based upon the dialog response and updated conversation state received from the second natural language processor; and
displaying, by the server computing device, the dialog response received from the second natural language processor on the user device as part of the first dialog flow.