Abstract: Systems and methods for associating data with a media asset are provided. A plurality of messages from a plurality of users is received at a remote server. Each message identifies a play position within the media asset at which the respective one of the users paused the media asset. A reference play position is selected based on the play position identified by one of the plurality of messages. A number representing how many of the messages identify a play position that corresponds to the reference play position is computed. In response to determining that the number exceeds a threshold, a frame of the media asset that corresponds to the reference play position is selected. Data associated with the media asset is updated based on the selected frame.

Abstract: Systems and methods for associating data with a media asset are provided. A plurality of messages from a plurality of users is received at a remote server. Each message identifies a play position within the media asset at which the respective one of the users paused the media asset. A reference play position is selected based on the play position identified by one of the plurality of messages. A number representing how many of the messages identify a play position that corresponds to the reference play position is computed. In response to determining that the number exceeds a threshold, a frame of the media asset that corresponds to the reference play position is selected. Data associated with the media asset is updated based on the selected frame.

Abstract: Systems and methods for generating recommendations are provided. User input that pauses a media asset at a play position may be received. Objects in a frame of the media asset corresponding the play position at which the user paused the media asset are identified. A determination is made as to which of the objects in the frame is associated with attributes that match a profile associated with the user. The object in the frame that is associated with attributes that match a profile associated with the user is automatically selected. A recommendation is generated for the user based on the selected one of the objects.