公开(公告)号：US20220310084A1

公开(公告)日：2022-09-29

申请号：US17211392

申请日：2021-03-24

Applicant: ADOBE INC.

Inventor： OLIVER BRDICZKA ,  Kyoung Tak Kim ,  Charat Maheshwari

IPC: G10L15/22 ,  G10L15/18 ,  G10L15/06 ,  G06F16/242 ,  G06F16/2453 ,  G06F16/248

Abstract: The present disclosure describes systems and methods for extensible search, content, and dialog management. Embodiments of the present disclosure provide a dialog system with a trained intent recognition model (e.g., a deep learning model) to receive and understand a natural language query from a user. In cases where intent is not identified for a received query, the dialog system generates one or more candidate responses that may be refined (e.g., using human-in-the-loop curation) to generate a response. The intent recognition model may be updated (e.g., retrained) the accordingly. Upon receiving a subsequent query with similar intent, the dialog system may identify the intent using the updated intent recognition model.

公开(公告)号：US11948566B2

公开(公告)日：2024-04-02

申请号：US17211392

申请日：2021-03-24

Applicant: ADOBE INC.

Inventor： Oliver Brdiczka ,  Kyoung Tak Kim ,  Charat Maheshwari

IPC: G10L15/22 ,  G06F16/242 ,  G06F16/2453 ,  G06F16/248 ,  G10L15/06 ,  G10L15/18

CPC classification number: G10L15/22 ,  G06F16/243 ,  G06F16/24539 ,  G06F16/248 ,  G10L15/063 ,  G10L15/18 ,  G10L2015/227

Abstract: The present disclosure describes systems and methods for extensible search, content, and dialog management. Embodiments of the present disclosure provide a dialog system with a trained intent recognition model (e.g., a deep learning model) to receive and understand a natural language query from a user. In cases where intent is not identified for a received query, the dialog system generates one or more candidate responses that may be refined (e.g., using human-in-the-loop curation) to generate a response. The intent recognition model may be updated (e.g., retrained) the accordingly. Upon receiving a subsequent query with similar intent, the dialog system may identify the intent using the updated intent recognition model.

公开(公告)号：US10410247B2

公开(公告)日：2019-09-10

申请号：US14597597

申请日：2015-01-15

Applicant: Adobe Inc.

Inventor： Sanjeev Kumar Biswas ,  Charat Maheshwari

IPC: G06Q30/02

Abstract: Audio sampling and user profile generation of advertisement control techniques are described. In one or more implementations, in a digital environment in which advertisements are dynamically selected for output and consumption by a user, a method is utilized to generate a user profile by one or more computing devices that serves as a basis for this dynamic selection. The method includes generating the user profile based on monitored consumption of content by a user, the consumption monitored through capture of audio data output in an environment of a user and identification of the audio data through comparison of hash values calculated from the captured audio data. The method also includes determining suitability of one or more of a plurality of advertisements for output to and consumption by the user based on the generated user profile and controlling output of the one or more of the plurality of advertisements to the user based on the determined suitability.