A GMM-Based User Model for Knowledge Recommendation
Nian Yang, Guoxin Wang, Jia Hao, Yan Yan, Hairong Han · 2017
With the exponential increase of available information, the phenomenon of information overload has received extensive research attentions. Knowledge recommender system (KRS) is an efficient way to decrease information overload, and the user model is very critical for KRS. This paper proposes a method to establish a user model based on Gaussian Mixture Model (GMM). In detail, we first select the keywords from knowledge databases, and then represent knowledge items with Vector Space Model (VSM). Next, for a certain user, the VSM of all scanned knowledge items and related scores rated by the user are combined together to be a new matrix, named as Vector Space Model with Rating(VSMR,with dimension of m times n), where the first n-1 columns represent the VSM of the items, and the final column lists the scores given by the user. And then the GMM-based user model is trained with VSMR. Finally, the trained user model is used to predict the user's ratings on the knowledge items and the items with the higher score are considered as user's interest, which will be recommended to the user. The proposed method is validated by two experiments, which indicate that the method works well.