USER-BASED RECOMMENDATION ENGINE FOR RECOMMENDING A HIGHLY-ASSOCIATED ITEM

摘要

The present invention relates to a recommendation engine. The recommendation engine searches for at least one item to be recommended which is associated with a reference item selected by an inquirer. The recommending engine includes: a query statement generating module for storing a plurality of item vectors as a plurality of documents, searching the plurality of documents for a reference document associated with the reference item, extracting a reference item vector, and if successful, generating a query statement including at least one user most highly associated with the reference item, wherein each of the plurality of item vectors consists of an element of a user-preference pair; and a search module for calculating correlations among each of the plurality of item vectors and providing at least one item to be recommended. Therefore, the disclosed technique can quickly search for items to be recommended.