Search systems based on user relevance and revenue generation

A system, comprising:

• one or more memories configured to store a query from a user, information about a user account of the user, and an item record of the item for a list of items, a parameter alpha, and a parameter beta;
• and
• at least one processor each coupled to at least one of the one or more memories and configured to perform operations comprising:
• generating an ordered relevance list of items, and an array of relevance scores corresponding to the ordered relevance list of items, wherein the array of relevance scores includes a relevance score of the item determined based on the item record of the item, the query from the user, and the information about the user account of the user;
• determining an initial reward value for an initial list of recommended items that is the ordered relevance list of items based on the array of relevance scores, and an array of revenue values including a revenue value of the item determined by the item record of the item, and the initial reward value is determined based on the parameter alpha assigned to the array of relevance scores to represent a first weight of the array of relevance scores, and the parameter beta assigned to the array of revenue values to represent a second weight of the array of revenue values;
• generating a next list of recommended items from the initial list of recommended items by switching positions of two or more items of the initial list of recommended items;
• calculating a next reward value associated with the next list of recommended items based on the parameter alpha, the parameter beta, a next array of relevance scores, and a next array of revenue values corresponding to the next list of recommended items;
• determining the next list of recommended items to be a list of recommended items in response to the query when an exit condition is met based on a comparison of the initial reward value and the next reward value;
• and
• presenting the list of recommended items to the user.

A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least a computing device, cause the computing device to perform operations comprising:

• receiving a query from a user;
• generating, by at least one computer processor, an ordered relevance list of items, and an array of relevance scores corresponding to the ordered relevance list of items, wherein the array of relevance scores includes a relevance score of an item determined based on an item record of the item, the query from the user, and information about a user account of the user;
• determining an initial reward value for an initial list of recommended items that is the ordered relevance list of items based on the array of relevance scores, and an array of revenue values including a revenue value of the item determined by the item record of the item, and the initial reward value is determined based on a parameter alpha assigned to the array of relevance scores to represent a first weight of the array of relevance scores, and a parameter beta assigned to the array of revenue values to represent a second weight of the array of revenue values;
• generating a next list of recommended items from the initial list of recommended items by switching positions of two or more items of the initial list of recommended items;
• calculating a next reward value associated with the next list of recommended items based on the parameter alpha, the parameter beta, a next array of relevance scores, and a next array of revenue values corresponding to the next list of recommended items;
• determining the next list of recommended items to be a list of recommended items in response to the query when an exit condition is met based on a comparison of the initial reward value and the next reward value;
• and
• presenting the list of recommended items to the user.