公开(公告)号：US20200380418A1

公开(公告)日：2020-12-03

申请号：US16995149

申请日：2020-08-17

Applicant: Google LLC

Inventor： Brian Strope ,  Yun-hsuan Sung ,  Wangqing Yuan

IPC: G06N20/00 ,  G06F16/35 ,  G06F16/332 ,  G06F16/33 ,  G06N5/04

Abstract: Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a “semantic textual similarity” task for which the encoder model can be used.

公开(公告)号：US20200184158A1

公开(公告)日：2020-06-11

申请号：US16792572

申请日：2020-02-17

Applicant: Google LLC

Inventor： James Kuczmarski ,  Vibhor Jain ,  Amarnag Subramanya ,  Nimesh Ranjan ,  Melvin Jose Johnson Premkumar ,  Vladimir Vuskovic ,  Luna Dai ,  Daisuke Ikeda ,  Nihal Sandeep Balani ,  Jinna Lei ,  Mengmeng Niu ,  Hongjie Chai ,  Wangqing Yuan

IPC: G06F40/58 ,  G06F40/47 ,  G06F16/33 ,  G06F16/332 ,  H04L12/58 ,  G06N20/00 ,  G06K9/62

Abstract: Techniques described herein relate to facilitating end-to-end multilingual communications with automated assistants. In various implementations, speech recognition output may be generated based on voice input in a first language. A first language intent may be identified based on the speech recognition output and fulfilled in order to generate a first natural language output candidate in the first language. At least part of the speech recognition output may be translated to a second language to generate an at least partial translation, which may then be used to identify a second language intent that is fulfilled to generate a second natural language output candidate in the second language. Scores may be determined for the first and second natural language output candidates, and based on the scores, a natural language output may be selected for presentation.

公开(公告)号：US10157615B2

公开(公告)日：2018-12-18

申请号：US15915599

申请日：2018-03-08

Applicant: Google LLC

Inventor： Bryan Horling ,  David Kogan ,  Maryam Garrett ,  Daniel Kunkle ,  Wan Fen Nicole Quah ,  Ruijie He ,  Wangqing Yuan ,  Wei Chen ,  Michael Itz

IPC: G10L21/00 ,  G10L15/22 ,  G06F17/30 ,  G10L15/18 ,  G10L15/30 ,  G10L15/26

Abstract: Techniques are described herein for chatbots to achieve greater social grace by tracking users' states and providing corresponding dialog. In various implementations, input may be received from a user at a client device operating a chatbot, e.g., during a first session between the user and the chatbot. The input may be semantically processed to determine a state expressed by the user to the chatbot. An indication of the state expressed by the user may be stored in memory for future use by the chatbot. It may then be determined, e.g., by the chatbot based on various signals, that a second session between the user and the chatbot is underway. In various implementations, as part of the second session, the chatbot may output a statement formed from a plurality of candidate words, phrases, and/or statements based on the stored indication of the state expressed by the user.

公开(公告)号：US20180197542A1

公开(公告)日：2018-07-12

申请号：US15915599

申请日：2018-03-08

Applicant: Google LLC

Inventor： Bryan Horling ,  David Kogan ,  Maryam Garrett ,  Daniel Kunkle ,  Wan Fen Nicole Quah ,  Ruijie He ,  Wangqing Yuan ,  Wei Chen ,  Michael Itz

IPC: G10L15/22 ,  G10L15/30 ,  G10L15/18 ,  G10L15/26

CPC classification number: G10L15/22 ,  G06F17/30867 ,  G10L15/1815 ,  G10L15/26 ,  G10L15/30

Abstract: Techniques are described herein for chatbots to achieve greater social grace by tracking users' states and providing corresponding dialog. In various implementations, input may be received from a user at a client device operating a chatbot, e.g., during a first session between the user and the chatbot. The input may be semantically processed to determine a state expressed by the user to the chatbot. An indication of the state expressed by the user may be stored in memory for future use by the chatbot. It may then be determined, e.g., by the chatbot based on various signals, that a second session between the user and the chatbot is underway. In various implementations, as part of the second session, the chatbot may output a statement formed from a plurality of candidate words, phrases, and/or statements based on the stored indication of the state expressed by the user.

公开(公告)号：US20240304184A1

公开(公告)日：2024-09-12

申请号：US18120216

申请日：2023-03-10

Applicant: GOOGLE LLC

Inventor： Roberto Pieraccini ,  Wangqing Yuan ,  Martin Baeuml

IPC: G10L15/197 ,  G06F40/35 ,  G10L15/06 ,  G10L15/18 ,  G10L15/22 ,  G10L15/30

CPC classification number: G10L15/197 ,  G06F40/35 ,  G10L15/063 ,  G10L15/1807 ,  G10L15/1815 ,  G10L15/22 ,  G10L15/30

Abstract: As part of an ongoing dialog between a user and an automated assistant, processor(s) can receive a natural language (NL) based input from the user during a turn of the ongoing dialog, obtain style signal(s) for the turn, and determine, based on the style signal(s), a NL based response style that is not specified in the NL based input. Further, the processor(s) can process, using a large language model (LLM), the NL based input and a NL based response style tag for the NL based response style to generate LLM output, determine, based on the LLM output, a NL based response in the NL based response style, and cause the NL based response to be rendered. In some implementations, a LLM behavior controller is utilized to determine the NL based response style, whereas in other implementations, the LLM is fine-tuned to determine the NL based response style.

公开(公告)号：US20240256533A1

公开(公告)日：2024-08-01

申请号：US18103973

申请日：2023-01-31

Applicant: GOOGLE LLC

Inventor： Wangqing Yuan ,  Bryan Christopher Horling

IPC: G06F16/2452 ,  G06F40/30

CPC classification number: G06F16/24522 ,  G06F40/30

Abstract: Techniques disclosed herein are directed towards generating structured data output based on processing a natural language user query using a semantic parser model. Many implementations include identifying one or more argument spans in the given natural language user query based on comparing an embedding space representation of a candidate argument with an embedding space representation of an example query, where the example query is provided by a developer. Various implementations include hotfixing an under-triggering model and/or an over-triggering model based on additional or alternative example queries provided by a developer.

公开(公告)号：US20240062111A1

公开(公告)日：2024-02-22

申请号：US18386015

申请日：2023-11-01

Applicant: GOOGLE LLC

Inventor： Brian Strope ,  Yun-Hsuan Sung ,  Wangqing Yuan

IPC: G06N20/00 ,  G06F16/33 ,  G06F16/332 ,  G06F16/35 ,  G06N5/04

CPC classification number: G06N20/00 ,  G06F16/3329 ,  G06F16/3344 ,  G06F16/3346 ,  G06F16/35 ,  G06N5/04

Abstract: Systems, methods, and computer readable media related to: training an encoder model that can be utilized to determine semantic similarity of a natural language textual string to each of one or more additional natural language textual strings (directly and/or indirectly); and/or using a trained encoder model to determine one or more responsive actions to perform in response to a natural language query. The encoder model is a machine learning model, such as a neural network model. In some implementations of training the encoder model, the encoder model is trained as part of a larger network architecture trained based on one or more tasks that are distinct from a “semantic textual similarity” task for which the encoder model can be used.

公开(公告)号：US20220405471A1

公开(公告)日：2022-12-22

申请号：US17351861

申请日：2021-06-18

Applicant: GOOGLE LLC

Inventor： David Kogan ,  Wangqing Yuan ,  Guanglei Wang ,  Vincent Lacey ,  Wei Chen ,  Shaun Post

IPC: G06F40/263 ,  G06F40/279 ,  G10L15/19 ,  G06F3/0482

Abstract: Implementations relate to determining a secondary language proficiency measure, for a user in a secondary language (i.e., a language other than a primary language specified for the user), where determining the secondary language proficiency measure is based on past interactions of the user that are related to the secondary language. Those implementations further relate to utilizing the determined secondary language proficiency measure to increase efficiency of user interaction(s), such as interaction(s) with a language learning application and/or an automated assistant. Some of those implementations utilize the secondary language proficiency measure in automatically setting value(s), biasing automatic speech recognition, and/or determining how to render natural language output.

公开(公告)号：US11322143B2

公开(公告)日：2022-05-03

申请号：US16712481

申请日：2019-12-12

Applicant: Google LLC

Inventor： Bryan Horling ,  David Kogan ,  Maryam Garrett ,  Daniel Kunkle ,  Wan Fen Nicole Quah ,  Ruijie He ,  Wangqing Yuan ,  Wei Chen ,  Michael Itz

IPC: G10L15/22 ,  G06F16/9535 ,  G10L15/18 ,  G10L15/30 ,  G10L15/26

Abstract: Techniques are described herein for chatbots to achieve greater social grace by tracking users' states and providing corresponding dialog. In various implementations, input may be received from a user at a client device operating a chatbot, e.g., during a first session between the user and the chatbot. The input may be semantically processed to determine a state expressed by the user to the chatbot. An indication of the state expressed by the user may be stored in memory for future use by the chatbot. It may then be determined, e.g., by the chatbot based on various signals, that a second session between the user and the chatbot is underway. In various implementations, as part of the second session, the chatbot may output a statement formed from a plurality of candidate words, phrases, and/or statements based on the stored indication of the state expressed by the user.

公开(公告)号：US20220027377A1

公开(公告)日：2022-01-27

申请号：US17497264

申请日：2021-10-08

Applicant: Google LLC

Inventor： David Kogan ,  Wangqing Yuan ,  Wei Chen ,  Bryan Horling ,  Michael Itz

IPC: G06F16/2457 ,  G06N20/00 ,  G06F16/332 ,  G06F40/169 ,  G10L15/22 ,  H04L12/58 ,  H04W4/14

Abstract: Methods, apparatus, systems, and computer-readable media are provided for automatically augmenting message exchange threads based on a detected tone of messages exchanged between participants. In various implementations, a message contributed to a message exchange thread involving one or more message exchange clients by a participant may be determined. In various implementations, an idle chatter score associated with the message may be calculated. In various implementations, either a conversational response to the message or content responsive to a search query generated based on the message may be selectively incorporated into the message exchange thread based at least in part on the idle chatter score. In some implementations, a search query suitability score associated with the message may also be calculated.