US 2022/0188133 A1Published application

Facilitating user device and/or agent device actions during a communication session

AssigneeGoogle Inc.

InventorsRobin Dua, Andrew Tomkins, Sujith Ravi

Life of this patent

Estimating the term…

Loading patent…

Abstract

Implementations are directed to facilitating user device and/or agent device actions during a communication session. An interactive communications system provides outputs, as outlined below, that are tailored to enhance the functionality of the communication session, reduce the number of dialog “turns” of the communications session and/or the number of user inputs to devices involved in the session, and/or otherwise mitigate consumption of network and/or hardware resources during the communication session. In various implementations, the communication session involves user device(s) of a user, agent device(s) of an agent, and the interactive communications system. The interactive communications system can analyze various communications from the user device(s) and/or agent device(s) during a communication session in which the user (via the user device(s)) directs various communications to the agent, and in which the agent (via the agent device(s)) optionally directs various communications to the user. The interactive communications system provides action performance element(s) and/or other output(s) that are each specific to a corresponding current intent and corresponding current action of the communication session.

Claims19

Dependent claims

A method implemented by one or more processors, the method comprising:

receiving one or more instances of natural language input during a communication session that includes a user computing device and an interactive communications system implemented by one or more of the processors, the instances of the natural language input including free-form input formulated by a user of the user computing device via a user interface input device of the user computing device;
generating a confidence score for a current intent of the communication session, wherein generating the confidence score is based on processing the natural language input using one or more intent models, the one or more intent models each being a corresponding trained machine learning model;
selecting the current intent based on the confidence score;
determining an action for the selected current intent;
resolving an agent for the action;
determining whether one or more criteria are satisfied for transmitting, to the user computing device, an action performance element that is based on the action;
in response to determining that the one or more criteria are satisfied:
transmitting, to the user computing device without first prompting any agent computing device of the agent, the action performance element that is based on the action;
and
in response to determining that the one or more criteria are not satisfied:
transmitting, to an agent computing device of the agent, a prompt related to the action, wherein the agent computing device is distinct from the user computing device, and
then transmitting the action performance element to the user computing device only if affirmative input is received from the agent computing device in response to transmitting the prompt.

9 more elements

A method implemented by one or more processors, the method comprising:

receiving one or more instances of natural language input during a communication session that includes a user computing device and an interactive communications system implemented by one or more of the processors, the instances of the natural language inputs including free-form input formulated by a user of the user computing device via a user interface input device of the user computing device;
processing the natural language input to generate annotations of the natural language input;
selecting, from a set of candidate intents, at least one current intent of the communication session, wherein selecting the current intent is based on processing the natural language input and the annotations using one or more intent models, the one or more intent models each being a corresponding trained machine learning model;
determining an action for the selected current intent;
resolving an agent for the action;
determining that an agent specific parameter, that is specific to the resolved agent, is not available for the action;
in response to determining that the agent specific parameter is not available for the action:
determining a stored parameter for a domain of the agent, the domain being a domain that encompasses the agent and a plurality of additional agents;
and
automatically initiating performance of the action with the stored parameter for the domain of the agent.

7 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/687,316Filed 4 Mar 2022
Priority: 22 Aug 2017
Published: 16 Jun 2022
Term ends: Not estimatedThe source records don’t support an estimate.

Classification13 codes

G06F 9/453
Execution arrangements for user interfaces: help systems
G06F 3/0481
Interaction techniques based on graphical user interfaces [GUI]: based on specific properties of the displayed interaction object or a…
G06F 3/167
Sound input; Sound output: audio in a user interface, e.g. using voice commands for navigating, audio feedback
G06F 40/20
Handling natural language data: natural language analysis

Family

Application62/548,928Us provisional application · filed 22 Aug 2017
Application16/621,769Continuation · filed 12 Dec 2019
Application17/687,316Continuation
US 11,269,666 B2Related patent
US 11,941,420 B2Related patent
US 2020/0143114 A1Published application
US 2022/0188133 A1This patent

From the US filing history. Foreign family members aren’t listed.