Interaction-based identification of items in content
AssigneeAmazon Technologies, Inc.
InventorsCharles Shearer Dorner, Robert Yuji Haitani, Jongwoo Lee
Life of this patent
Estimating the term…
Abstract
Features are provided for the identification of objects in content, such as video content, by analyzing data regarding viewer interactions with other content. The interaction data can be analyzed to identify correlations between time periods of the video (or the video in general) and items that are the subject of, or otherwise associated with, other content with which viewers interact. These correlations can be used to provide, to subsequent viewers of the video, information regarding the items shown in the video, recommendations regarding items or content that may be of interest to subsequent viewers of the video, etc.
Claims20
A system comprising:
- a non-transitory computer-readable memory storing executable instructions;
- and
- one or more computing devices in communication with the non-transitory computer-readable memory, the one or more computing devices programmed by the executable instructions to at least:
- obtain data regarding interactions of a plurality of viewers with content associated with an item, wherein the interactions comprise at least one of:
- browsing the content associated the item, performing a search for the content associated with the item, or initiating a purchase of the item using the content associated with the item, and wherein interactions of individual viewers of the plurality of viewers are performed within a threshold amount of time of accessing a video;
- identify an association between at least a portion of the interactions and at least a portion of the video based at least partly on an analysis of times at which the interactions are performed and times at which the video is accessed;
- analyze the portion of the video for a visual representation of the item;
- determine that the visual representation of the item is present in the portion of the video;
- provide the video to a computing device associated with an additional viewer, wherein the additional viewer is not in the plurality of viewers;
- and
- provide information regarding the item to the computing device associated with the additional viewer based at least partly on the visual representation of the item being present in the portion of the video, wherein the computing device associated with the additional viewer is configured to display the information regarding the item in connection with displaying the portion of the video.
A computer-implemented method comprising:
- as performed by a content management system comprising a server computing device configured to execute specific instructions,
- analyzing data regarding timing of actions associated with an item relative to timing of viewing events associated with a video, wherein the actions associated with the item comprise at least one of:
- browsing content regarding the item, performing a search related to the item, or purchasing the item;
- identifying an association between at least a portion of the actions and at least a portion of the video;
- determining that the video likely comprises a visual representation of the item based at least partly on the association between the portion of the actions and the portion of the video;
- and
- providing, to a user computing device, information regarding the item based at least partly on a request to access the video.
A non-transitory computer storage medium storing executable code, wherein the executable code configures a computing system to perform a process comprising:
- analyzing data regarding timing of actions associated with an item relative to timing of events associated with consuming a content item, wherein the actions associated with the item comprise at least one of:
- browsing content regarding the item, performing a search related to the item, or purchasing the item;
- identifying an association between at least a portion of the actions and at least a portion of the content item;
- determining that the content item likely references the item based at least partly on the association between the portion of the actions and the portion of the content item;
- and
- providing, to a user computing device, information regarding the item based at least partly on a request to access the content item.
Dependencies are inferred from the claim text.