Combining parameters of multiple search queries that share a line of inquiryselectively storing, with multiple user accounts and/or to a shared assistant device: speech recognition biasing, nlu biasing, and/or other data

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

• receiving, via one or more microphones of a shared assistant device that is associated with multiple user accounts, audio data that captures a spoken utterance of a first user;
• generating, based on processing the audio data, a transcription that corresponds to the spoken utterance captured in the audio data;
• determining, based on processing the transcription that corresponds to the spoken utterance, at least one assistant action to perform in response to receiving the spoken utterance;
• determining one or more biasing parameters that are based on the spoken utterance, the one or more biasing parameters including:
• one or more speech biasing parameters that are based on one or more terms of the transcription;
• and/or
• one or more natural language understanding biasing parameters that are based on the at least one assistant action;
• determining whether to store the one or more biasing parameters in association with the shared assistant device;
• in response to determining to store the one or more biasing parameters in association with the shared assistant device:
• storing the one or more biasing parameters in association with the shared assistant device,
• wherein storing the one or more biasing parameters in association with the shared assistant device causes future spoken utterances from at least all authorized users to be processed using the one or more biasing parameters.

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

• receiving, via one or more microphones of a shared assistant device that is associated with multiple user accounts, audio data that captures a spoken utterance of a first user;
• generating, based on processing the audio data, a transcription that corresponds to the spoken utterance captured in the audio data;
• determining, based on processing the transcription that corresponds to the spoken utterance, at least one assistant action to perform in response to receiving the spoken utterance, the at least one assistant action comprising storing an association of first data based on one or more first terms of the transcription to second data based on one or more second terms of the transcription;
• classifying the spoken utterance as being spoken by a first user account corresponding to the first user, the first user account being one of the multiple user accounts associated with the shared assistant device;
• determining whether to store the association of the first data to the second data as accessible to only the first user account or as accessible to the first user account and to one or more additional user accounts of the multiple user accounts, each of the additional user accounts being in addition to the first user account and being associated with a corresponding user that is in addition to the first user;
• in response to determining to store the association as accessible to only the first user account:
• storing the association accessible to only the first user account without enabling access to the association by any of the additional user accounts;
• in response to determining to store the association as accessible to the first user account and to the one or more additional user accounts:
• storing the association as accessible to the first user account and as accessible by the one or more of the additional user accounts.