Computer system and graphical user interface for determining diamond and ring attributes using a predictive computer model
Ring attributes may be predicted using a predictive model. A diamond selection system determines predicted ring attributes by obtaining and analyzing data related to a recipient, such as social media activity. The diamond selection system may determine search parameters based on predicted ring attributes and present a curated selection or list of candidate stones and/or settings that match the determined search parameters, for example, in a user interface of a mobile electronic device.
1. A computer-implemented method for automatically determining search parameters for a jewelry selection system having an interactive search interface, the computer-implemented method performed using one or more processors of a computer system, the computer-implemented method comprising:
causing display of, on a personal mobile device, a user interface for selecting a jewelry product;
receiving, via the user interface, a username for an account of a user of a social media network, the account associated with an image set accessible using the username;
in response to authenticating the account using the username, performing computer vision object recognition to determine a set of user classification labels associated with the image set;
determining a set of predicted jewelry attributes by providing the set of user classification labels to a predictive model;
determining search parameters for the interactive search interface based on the set of predicted jewelry attributes; and
causing display of a jewelry listing based on the search parameters.
2. The method of claim 1 , wherein determining the set of predicted jewelry attributes comprises determining a matching group of training data sets by determining a measure of similarity between the set of user classification labels and classification labels of one or more training data sets stored by the jewelry selection system.
3. The method of claim 2 , wherein determining the set of predicted jewelry attributes comprises determining at least one of a probability distribution or a mathematical average of a jewelry attribute among multiple training data sets of the matching group of training data sets.
4. The method of claim 1 , wherein the search parameters comprise a diamond shape and one or more search categories, each search category corresponding to a diamond feature graded along a scale.
5. The method of claim 1 , further comprising:
determining, based on the search parameters, a set of jewelry listings for presentation in the user interface;
determining a presentation order of a set of jewelry listings based on the search parameters; and
displaying at least a portion of the set of jewelry listings in the user interface.
6. The method of claim 5 , wherein:
the set of jewelry listings is a first set of jewelry listings; and
the presentation order is a first presentation order; and
the method further comprises:
receiving a user input indicating a preference regarding a listing of the first set of jewelry listings;
updating the search parameters based on the preference;
determining a second set of jewelry listings based on the updated search parameters; and
determining a second presentation order for the second set based on the updated search parameters.