SIMILARITY METRIC RELATIVIZED TO A USER'S PREFERENCES

摘要

Mathematical technologies for recommending content to a user based on a user's preferences are disclosed. Embodiments of these technologies can generate a probabilistic representation of a data set, and then adjust the probabilistic representation to reflect a user-specific weighting scheme. The user preference-adjusted representation of the data set can be used to recommend content to the user.

主权项

1. A method for recommending content to a user based on a data set of the user, the method comprising, by a computing system comprising one or more computing devices:
creating a probability distribution of the data set, the data set comprising a plurality of items, each item comprising one or more occurrences of a plurality of features, the probability distribution comprising, for each item and feature pair in the data set, a probability value indicative of the proportion of total occurrences of any feature accounted for by that item and feature pair; receiving one or more input terms in response to one or more inputs entered by a user; accessing a user-preference feature distribution indicative of one or more preferences of the user, the user-preference feature distribution indicating a likelihood that the user will select at least one of the pluralities of features or items relative to all other ones of the plurality of features or items of the data set; determining a user-specific similarity metric indicative of a similarity between the one or more input terms and the items of the data set based on the probability distribution of the data set and the user-preference feature distribution; and outputting a content recommendation based on the user-specific similarity metric.