METHODS AND SYSTEMS FOR ACTIVITY-BASED RECOMMENDATIONS

摘要

Embodiments of computer-implemented methods and systems for activity-based recommendations are described. One example embodiment includes receiving data indicating historical activities of a user community, the historical activities including historical activities of the target user, selecting a reference group of users from the user community based on analysis of the historical activities of the target user, receiving generally current time activities of the reference group of users, the generally current time activities including those activities that have occurred within a defined time window, and recommending items to the target user based on the generally current time activities of the reference group of users.

主权项

1. A system comprising:
a reception module comprising, one or more processors of a machine and configured to receive real-time action from a target user, the real-time action is not an action within an activity network of the target user; a cluster module comprising at least a first processor among the one or more processors of the machine and configured to re-categorize activity data indicating actions, including the real-time actions engaged by multiple users, into updated activity networks by performing an analysis of the activity data; an association module comprising at least a second processor among the one or more processors of the machine and configured to re-determine a subset of the multiple users that engaged in one or more actions categorized into the updated activity networks that includes the real-time action engaged by the target user; and a recommendation module comprising at least a third processor among the one or more processors of the machine and configured to rank a plurality of recommendations directed at the target user used on a count of activity signals received from the multiple users of the activity network, the activity signals include real-time actions from the multiple users, and display the real-time ranked recommendations.