Human knowledge is inherently complex, involving diverse and differentiated components that are woven together by multiplex, functional ties. It is also highly dynamic, manifesting incremental growth punctuated by unexpected breakthroughs and transformations. Although the complex nature of knowledge production has been long studied, it is only recently that systematic and quantitative understandings are made possible by the availability of digital traces of knowledge products, the development of machine learning techniques, and the advancement of network science.

Example research includes drug discovery using semantic networks between chemicals and diseases, intelligent systems based on knowledge graphs extracted from Wikipedia, and prediction of future technologies from networks of patent classes, just to name a few. And the focus is shifting from citation patterns to semantic or functional relationships between knowledge elements in those systems, whether they being chemicals, proteins, genes, technologies, concepts or simply words.

There is now a wide spectrum of network models and techniques that are developed by researchers on knowledge systems from different perspectives. This symposium will bring together researchers from a wide range of fields working on domain-specific or general knowledge networks, and survey state-of-the-art theories, models, methods, and findings in studies of the complex systems of human knowledge, with the goal to identify regularities and commonalities in different knowledge networks, share cutting-edge methods and discoveries, foster collaborations across knowledge domains, and eventually, generate new knowledge from old.