中文摘要：

小世界网络在聚类应用中具有良好的性质，贝叶斯网络在概率推理中也得到了广泛的研究．将小世界网络和贝叶斯网络结合起来，形成了一种混合推荐模型．该混合模型由两层组成，分别是用户层和商品层．其中小世界网络用于描述用户层内用户-用户结点间的关系，贝叶斯网络用于描述商品层内商品-商品结点，以及层间用户-商品结点间的偏好关系．对小世界网络的用户聚类方法、贝叶斯网络结构和参数学习方法、以及两层混合模型的椎荐算法进行了描述，实验表明，该模型能够很好地表示用户-用户、商品-商品、以及用户-商品间的关系，推荐结果具有良好的准确度．

英文摘要：

A hybrid model for personalized recommendation that is based on small world network and Bayesian network is presented. Small world network has a good property in clustering and Bayesian network is compatible for probability inference. The hybrid model consists of two layers. One is for consumers＇layer and the other is for produce＇s layer. The relationships among nodes of consumers are described by small world network at low layer. The implications among nodes of produce are represented by Bayesian network at high layer. Directed arcs denote the tendency between consumer~ layer and produce~ layer. We also introduce several algorithms for clustering based on small world network, structure learning and parameter learning and recommended based this model. The experimentation shows that the model can well represent the relationships between consumer to consumer, produce to produce and consumer to produce. The result of recommendation based this hybrid mode is better than other.