Multi-strategy generation of product recommendations

A computer-implemented method for dynamically generating and providing recommendations to a user, the method comprising:

• retrieving information that reflects previous interactions of a plurality of users with a plurality of items available from one or more retailers;
• receiving information that indicates a selection by a first user of a subset of one or more of the plurality of items;
• and
• under control of one or more configured computing systems,
• automatically determining a first recommendation group of multiple of the plurality of items to recommend to the first user using a first recommendation strategy that is based at least in part on the selected subset of the one or more items, and determining for each of the multiple items of the first group an associated relevance level;
• generating a Web page for use in displaying the multiple items of the first group to the first user in such a manner that items of the first group with higher associated relevance levels are displayed more prominently than other items of the first group with lower associated relevance levels;
• after the generated Web page is displayed to the first user and after receiving an indication of one of the multiple items of the first group that the first user has selected, automatically determining a second recommendation group of multiple of the plurality of items to recommend to the first user that is distinct from the first group, the determining of the second group using a combination of multiple recommendation strategies that are based at least in part on the selected subset of the one or more items and on the retrieved information about the previous interactions of the plurality of users and on the selected one item of the first group, and determining for each of the multiple items of the second group an associated relevance level;
• and
• sending information about the determined second group of multiple items for display to the first user as part of a modification of the displayed generated Web page that is performed without reloading the generated Web page, such that items of the second group with higher associated relevance levels are displayed more prominently than other items of the second group with lower associated relevance levels.

A computer-readable medium whose contents configure a computing device to generate and provide recommendations to a user, by performing a method comprising:

• retrieving information that reflects interactions of a plurality of users with a plurality of items;
• automatically determining a first group of multiple of the plurality of items to recommend to a first user using a first collection of one or more of multiple recommendation strategies, the first group of multiple items being determined based on at least one of one or more items previously selected by the first user and of the information about the interactions of the plurality of users with the plurality of items;
• providing information for use as part of a presentation to the first user to indicate the multiple items of the first group;
• after receiving an indication of one of the multiple items of the first group that the first user has selected, automatically determining a second group of multiple of the plurality of items to recommend to the first user using a second collection of one or more of the multiple recommendation strategies that is distinct from the first collection, the second group of multiple items being determined based on multiple of the one or more items previously selected by the first user and of the selected one item of the first group and of the information about the interactions of the plurality of users with the plurality of items;
• and
• providing information for use as part of dynamically modifying the presentation to the first user to indicate the multiple items of the second group.

A computing system configured to generate and provide recommendations to users, comprising:

• one or more processors;
• and
• a product recommendation system configured to, when executed by at least one of the one or more processors, provide recommendations to multiple users by, for each of the users:
• obtaining information that reflects interactions of a plurality of users with a plurality of items;
• automatically determining a first group of multiple of the plurality of items to recommend to the user using one or more recommendation strategies, the first group of multiple items being determined based on at least one of one or more items previously selected by the user and of the information about the interactions of the plurality of users with the plurality of items;
• providing information for use as part of a display to the user to indicate the multiple items of the first group;
• after receiving an indication of at least one of the multiple items of the first group that the user has selected, automatically determining a second group of multiple of the plurality of items to recommend to the user using a combination of multiple recommendation strategies, the second group of multiple items being distinct from the first group of multiple items and being determined based on multiple of the one or more items previously selected by the user and of the selected at least one items of the first group and of the information about the interactions of the plurality of users with the plurality of items;
• and
• providing information for use as part of dynamically modifying the display to the user to indicate the multiple items of the second group.