中文摘要：

海量数字资源中蕴含着大量科研实体，从中可以识别、挖掘和发现大量的潜在科研关系，通过构建科研实体关系网络，可以推进海量信息资源的揭示、组织与深度聚合，有效地促进知识导航、智能检索、科学评价和专家识别等知识服务。文章在梳理国内外科研关系网络相关研究的基础上，分析了科研实体的语义描述、识别与抽取，以及科研实体关系揭示，评价与融合等关键问题，提出了构建科研关系网络的技术框架及方法路线，并分析其应用前景。

英文摘要：

There are myriads of research entities in the huge amount digital resources, and various potential research relations among the research entities can be detected and mined. Building research relation network can boost the revealation, organization and aggregation of the huge amount digital resources, and promote knowledge services, such as knowledge navigation, intelligent retrieval, scientific evaluation and expert identification. Based on the previous studies on research relation network, this paper analyzes the key issues including both semantic description, identification and extraction of research entities and revealation, evaluation and integration of research relations, and then proposes a technical framework and method to construct research relation network. And this paper also discusses the promising applications.