公开(公告)号：US20200372399A1

公开(公告)日：2020-11-26

申请号：US16419676

申请日：2019-05-22

Applicant: Adobe Inc.

Inventor： Tak Yeon Lee ,  Jonggi Hong ,  Eunyee Koh

IPC: G06N20/00 ,  G06F9/451 ,  G06F16/955

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for determining and resolving misalignments between digital messages containing links and corresponding external digital content. For example, in one or more embodiments, the disclosed systems extract a plurality of alignment classification features from a digital link in a digital message and corresponding external digital content. Based on the alignment classification features and using a machine learning classification model, the disclosed system can generate alignment probability scores for a plurality of misalignment classes. The disclosed system can report identified misalignments of corresponding misalignment classes in a misalignment identification user interface. Furthermore, the disclosed system can receive publisher input via the misalignment identification user interface to further personalize the machine learning classification model.

公开(公告)号：US11341204B2

公开(公告)日：2022-05-24

申请号：US16419676

申请日：2019-05-22

Applicant: Adobe Inc.

Inventor： Tak Yeon Lee ,  Jonggi Hong ,  Eunyee Koh

IPC: G06F16/955 ,  G06N20/00 ,  G06F9/451 ,  G06F16/93

Abstract: Methods, systems, and non-transitory computer readable storage media are disclosed for determining and resolving misalignments between digital messages containing links and corresponding external digital content. For example, in one or more embodiments, the disclosed systems extract a plurality of alignment classification features from a digital link in a digital message and corresponding external digital content. Based on the alignment classification features and using a machine learning classification model, the disclosed system can generate alignment probability scores for a plurality of misalignment classes. The disclosed system can report identified misalignments of corresponding misalignment classes in a misalignment identification user interface. Furthermore, the disclosed system can receive publisher input via the misalignment identification user interface to further personalize the machine learning classification model.