Nowadays some online research platforms (e.g., Web of Science or IEEE Xplore) provide bibliographic content and tools to access, analyze, and manage the world's leading journals and conference proceedings in sciences, social science, arts, and humanities. However, when facing increasingly mass literature, it’s very difficult for researchers to effectively and systematically acquire the knowledge structure about a particular topic by using traditional literature reviewing method. Therefore we need explore new knowledge discovery tools for knowledge representation in an effective and efficient way. This paper proposes a knowledge system meta-network model by identifying the concepts representing entities and relationships from bibliometric data, and a methodology framework for meta-network modeling and analysis by using integrated techniques, including text mining, network text analysis, social network analysis, longitudinal network analysis and visualization. Case study using the Web of Science database as data source, explores the knowledge structure and interdisciplinary cooperation mode, as well as hot topics evolution in the field of World Trade Web.