| CPC G06Q 30/0631 (2013.01) [G06F 3/017 (2013.01); G06Q 30/0629 (2013.01); G06Q 30/0643 (2013.01); G06V 40/20 (2022.01)] | 16 Claims |

|
1. A computer-implemented method for providing an item recommendation based on user gesture, the method comprising:
receiving, by one or more processors, a gesture from a user performed on a displayed first content item;
generating, by the one or more processors, a second display of the first content item comprising a marking of the gesture by the user;
determining, by the one or more processors, a feature of the item based on a location of the gesture on a display of the first content item;
determining a preference value for the feature based on a comparison of the gesture to predetermined gestures associated with different preference values;
determining, by the one or more processors and via a machine learning algorithm, an item preference of the user based on one or more of the gesture, the feature of the item, or the preference value;
identifying, by the one or more processors, an item recommendation based on the item preference of the user;
determining a similarity level of the item recommendation by comparing the item recommendation to the item preference of the user;
determining whether the similarity level of the item recommendation is equal to or exceeds a predetermined threshold; and
outputting, by the one or more processors, the item recommendation to the user, the output including a display of the item recommendation and an indicator of the similarity level when the similarity level of the item recommendation is equal to or exceeds the predetermined threshold.
|