动态商品名称体系的建立

在分布式电子市场中，要提供一个推荐商品命名和分类体系作为市场参与方用来交易与通信的标准，为此，提出了一种中心推荐聚类算法。该算法能在生成的聚类中，根据异质属性的影响因子，确立具有代表意义的推荐概念及商品名称间的层次关系。通过收集分布式电子市场中不同参与方对商品命名和分类的信息，使用中心推荐聚类算法动态建立推荐商品名称体系。其中，商品名称体系包括商品的推荐命名、名称间的关系以及其定义。实验结果表明，该算法是可行的。

Constructing Dynamic Recommended Product Ontologies

Recommended product ontologies are required for trading between buyers and sellers in distributed eMarketplace. Recommended product ontologies periodically built manually cannot include information of some new product names timely. A center recommendation clustering algorithm was provided. According to the values of heterogeneous attributes, the recommend product names can be selected at the clusters, which is produced by this algorithm. At the same time, this algorithm can create the hierarchical relations between the product names. It collects the definitions of product names of all participants in the distributed eMarketplace. Recommended product ontologies were built. These ontologies include the relations and definitions of product names, which come from different participants in the distributed eMarketplace. Finally, a case was illustrated by this method. The result shows that this method is feasible.