Abstract: 【Purpose/significance】To reveal the core disciplines and topics with multiple roles，and help researchers and management
departments get the overall cognition of interdisciplinary field, we construct an effective analysis framework of cognitive structure.
【Methods/process】Combining with the cognitive needs of the interdisciplinary field, an analysis framework of cognitive structure based on the "discipline-keyword" 2-mode network was designed. The framework was composed of three levels as knowledge units, knowl? edge relationships and knowledge communities, and two dimensions as disciplines and topics. A comparative analysis of heterogeneous networks and homogeneous networks following this framework was conducted to mine the cognitive characteristics of interdisciplinary field. Then, we performed the empirical analysis on the Covid-19 research field.【Result/conclusion】Compared with homogeneous net? works, the results show our framework has three advantages. Firstly, it helps to reveal the composite characteristics of knowledge units. Most high co-occurrence disciplines also have great knowledge contributions. The majority of core keywords in interdisciplinary field are also hot spots in multiple disciplines. Secondly, it excavates the connecting content of knowledge relationships. Common topics studied by discipline combinations and disciplines study keyword combinations could be further mined. Thirdly, it discovers the rela? tionship among knowledge communities and discipline communities. The discipline-keyword communities connect the discipline com? munities and co-word communities as the middle layer in the classification relationships.【Innovation/limitation】This research pro? vides method support for systematic mining cognitive structure and in-depth understanding of multiple characteristics of core content of interdisciplinary field. The community discovery methods and cognitive dimensions need to be further optimized.