The present invention relates to a method and
system for personalized recommendation based on
label information. The method comprises: generating a three-dimensional table according to the state of marking the identification of each product according to the information of each
label corresponding to the identification of each user, establishing a user product interaction
matrix model and a
product label relation
matrix model according to the three-dimensional table; constructing a joint
decomposition model corresponding to the user identification, the production identification and the
label information according to the user product interaction
matrix model and the
product label relation matrix model; employing a Bayes personalized
ranking method to solve the joint
decomposition model to obtain a plurality of parameter values; obtaining the preference degrees of the user identification to the identification of each product according to the parameter values; and selecting the recommendation identifications from the identifications of the products according to the preference degrees, and recommending the information of the products corresponding to the recommendation identifications to terminals whether the corresponding user identifications are located. Therefore, the method and
system for personalized recommendation based on the label information can solve the limitation of the data sparsity in the label information to improve the precision of the personalized
ranking so as to improve the accuracy of the recommendation.