INFORMATION RECOMMENDATION APPARATUS, INFORMATION RECOMMENDATION METHOD, AND INFORMATION RECOMMENDATION PROGRAM

摘要

PROBLEM TO BE SOLVED: To provide an information recommendation technology for recommending items across multiple domains with high accuracy.SOLUTION: An information recommendation apparatus for recommending an item for a user includes: partial matrix extraction means which refers to an item history database including a user ID, an item ID, and a domain ID, performs constrained clustering on an item by use of the domain ID information as a constraint, acquires an item group belonging to a cluster strongly associated with the user, and extracts a partial matrix on the basis of the item group; and recommended item determination means which creates a partial adjacency matrix from the partial matrix, calculates a degree of association between the user and the items by use of the partial adjacency matrix, and determines a recommended item on the basis of the calculated degree of association.