公开(公告)号：US10990759B1

公开(公告)日：2021-04-27

申请号：US16050488

申请日：2018-07-31

Applicant: Amazon Technologies, Inc.

Inventor： Mikhail Sosonkin ,  Michael Egan Butler ,  Ken F Eggers

IPC: G06F40/30 ,  G06Q10/06 ,  H04L9/32 ,  G06N20/00 ,  G06F16/2457

Abstract: This disclosure describes techniques implemented partly by a cloud-based service provider for providing a sentiment-analysis service to determine whether sentiments, or attitudes, of users have significantly decreased. The sentiment-analysis service is composed of various services provided by the cloud-based service in order to determine significant negative deviations in sentiment of users towards an entity. Further, the sentiment-analysis service utilizes various techniques, such as encryption, to privatize the process such that private user information is protected. In some examples, an entity may not know the content of communications or the reasons the user's sentiment has dropped, but may be alerted of a general negative change in sentiment towards the entity by the user.

公开(公告)号：US10740565B1

公开(公告)日：2020-08-11

申请号：US16050488

申请日：2018-07-31

Applicant: Amazon Technologies, Inc.

Inventor： Mikhail Sosonkin ,  Michael Egan Butler ,  Ken F Eggers

IPC: G06F40/30 ,  G06Q10/06 ,  H04L9/32 ,  G06N20/00 ,  G06F16/2457

Abstract: This disclosure describes techniques implemented partly by a cloud-based service provider for providing a sentiment-analysis service to determine whether sentiments, or attitudes, of users have significantly decreased. The sentiment-analysis service is composed of various services provided by the cloud-based service in order to determine significant negative deviations in sentiment of users towards an entity. Further, the sentiment-analysis service utilizes various techniques, such as encryption, to privatize the process such that private user information is protected. In some examples, an entity may not know the content of communications or the reasons the user's sentiment has dropped, but may be alerted of a general negative change in sentiment towards the entity by the user.