Processing complex utterances for natural language understanding

Patent metadata

Field	Value
Publication number	US-2023032575-A1
Application number	US-202217882874-A
Country	US
Kind code	A1
Filing date	Aug 8, 2022
Priority date	Sep 29, 2016
Publication date	Feb 2, 2023
Grant date	—

Abstract

A system capable of performing natural language understanding (NLU) on utterances including complex command structures such as sequential commands (e.g., multiple commands in a single utterance), conditional commands (e.g., commands that are only executed if a condition is satisfied), and/or repetitive commands (e.g., commands that are executed until a condition is satisfied). Audio data may be processed using automatic speech recognition (ASR) techniques to obtain text. The text may then be processed using machine learning models that are trained to parse text of incoming utterances. The models may identify complex utterance structures and may identify what command portions of an utterance go with what conditional statements. Machine learning models may also identify what data is needed to determine when the conditionals are true so the system may cause the commands to be executed (and stopped) at the appropriate times.

First claim

1 .- 20 . (canceled)

21 . A computer-implemented method, comprising:
determining that an input utterance includes a first portion corresponding to a command and a second portion corresponding to a condition; based at least in part on the command, causing at least one first component to transition from a first state to a second state; after the at least one first component has transitioned to the second state, determining that the condition is satisfied; and based at least in part on the condition being satisfied, causing the at least one first component to transition from the second state to the first state.

22 . The computer-implemented of claim 21 , further comprising:
receiving, from at least one second component, first data, wherein the determining that the condition is satisfied is based at least in part on the first data.

23 . The computer-implemented of claim 22 , further comprising:
determining that the condition corresponds to a first data type; and based at least on part on the condition corresponding to the first data type, determining to use the first data to determine satisfaction of the condition.

24 . The computer-implemented of claim 22 , wherein the at least one second component comprises a clock and the first data comprises time data.

25 . The computer-implemented of claim 22 , wherein the at least one second component comprises an appliance and the first data represents an operational state of the appliance.

26 . The computer-implemented of claim 22 , wherein the at least one second component comprises a thermometer and the first data comprises temperature data.

27 . The computer-implemented of claim 22 , wherein the at least one second component comprises a light bulb and the first data represents an operational state of the light bulb.

28 . The computer-implemented of claim 22 , wherein the at least one second component comprises a global positioning system (GPS) sensor and the first data comprises location data.

29 . The computer-implemented of claim 21 , wherein the at least one first component comprises a thermostat.

30 . The computer-implemented of claim 21 , wherein the at least one first component comprises a music player.

31 . The computer-implemented of claim 21 , wherein the at least one first component comprises a light bulb.

32 . A system, comprising:
at least one processor; and at least one computer-readable medium encoded with instructions that, when executed by the at least one processor, cause the system to:
determine that an input utterance includes a first portion corresponding to a command and a second portion corresponding to a condition;
based at least in part on the command, cause at least one first component to transition from a first state to a second state;
after the at least one first component has transitioned to the second state, determine that the condition is satisfied; and
based at least in part on the condition being satisfied, cause the at least one first component to transition from the second state to the first state.

33 . The system of claim 32 , wherein the at least one computer-readable medium is further encoded with additional instructions which, when executed by the at least one processor, further cause the system to:
receive, from at least one second component, first data; and determine that the condition is satisfied based at least in part on the first data.

34 . The system of claim 33 , wherein the at least one computer-readable medium is further encoded with additional instructions which, when executed by the at least one processor, further cause the system to:
determine that the condition corresponds to a first data type; and based at least on part on the condition corresponding to the first data type, determine to use the first data to determine satisfaction of the condition.

35 . The system of claim 33 , wherein the at least one second component comprises a clock and the first data comprises time data.

36 . The system of claim 33 , wherein the at least one second component comprises a household appliance and the first data represents an operational state of the household appliance.

37 . The system of claim 33 , wherein the at least one second component comprises a thermometer and the first data comprises temperature data.

38 . The system of claim 33 , wherein the at least one second component comprises a light bulb and the first data represents an operational state of the light bulb.

39 . The system of claim 33 , wherein the at least one second component comprises a global positioning system (GPS) sensor and the first data comprises location data.

40 . The system of claim 32 , wherein the at least one first component comprises a thermostat.

41 . The system of claim 32 , wherein the at least one first component comprises a music player.

42 . The system of claim 32 , wherein the at least one first component comprises a light bulb.

Assignees

• Amazon Tech Inc

Inventors

• Erbas Cengiz
• Kollar Thomas
• Sikka Avnish
• Matsoukas Spyridon
• Reavely Simon Peter

Classifications

• G10L15/1822

  Parsing for meaning understanding · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/063

  Training · CPC title

• G10L15/22 Primary

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L15/02

  Feature extraction for speech recognition; Selection of recognition unit · CPC title