中文摘要：

数据模型的研究是目前数据空间中研究的主要问题之一,是数据空间管理系统提供其它服务的基础.由于数据空间中多种异构性数据资源的共存和松散连接的特点,有必要提出一个有效、简单而且通用的数据模型来描述和管理这些数据.本文提出了分层的图数据模型lgDM来描述数据空间中的各种数据并捕捉实体间和实体类间的语义关联信息;并给出了实体关联关系挖掘的不同策略和对图加权重的方法.lgDM具有较好地通用性和扩展性,实验结果表明所提出模型的可行性和有效性.

英文摘要：

Research of data model is currently one of the hot research topics in dataspaces which is the foundation of other services in dataspaces management system.Due to co-existence and loosely connected of large heterogeneous data sources,it is necessary to propose an effective,simple and general data model to describe and manage all these data.This paper proposes a layered graph data model called lgDM to describe all kinds of data and to capture semantic associations among entities and entity classes in dataspaces.It also proposes different strategies of entity association mining and graph weighting method.lgDM is with generality and can be extended easily.Experiments results show feasibility and effectiveness of the proposed data model.