Linzhuo Li
Assistant Professor of Sociology, Zhejiang University
I use computational and qualitative methods to uncover hidden structures of knowledge and cultural complex systems, focusing on uncertainty and innovation.
Assistant Professor of Sociology, Zhejiang University
I use computational and qualitative methods to uncover hidden structures of knowledge and cultural complex systems, focusing on uncertainty and innovation.
I am a sociologist who uses both computational and qualitative methods to uncover hidden structures of knowledge and cultural complex systems.
I am particularly interested in uncertainty, recombination of knowledge, and disruptive changes in the knowledge economy. My work combines network analysis, machine learning, and qualitative methods to understand how new scientific knowledge is created, and how uncertainty is transformed in contemporary societies.
Key findings from my research on knowledge, innovation, and uncertainty
Using over 4.6 million online job recruitment records across 288 Chinese cities, this study reveals that explicit knowledge (measured by education requirements) and tacit knowledge (measured by experience requirements) exhibit opposite urban scaling patterns. Jobs demanding higher explicit knowledge concentrate more sharply in large cities, while jobs requiring deeper tacit know-how scale more slowly — suggesting that the engines of continued urban growth are increasingly biased towards institutionalized, academic knowledge rather than experience-based expertise.
Analyzing 30 manufacturing sub-sectors across 205 Chinese cities, this study disentangles urban scaling, returns to agglomeration, and knowledge complexity. A key finding challenges conventional wisdom: sectors that concentrate most in large cities are not those with the highest productivity returns, but those with the greatest knowledge complexity. This suggests that the growth of complex economic activities hinges more on urban diversity (the availability of diverse knowledge components for recombination) than on efficiency gains from scale, reframing how we understand the geography of innovation.
This study examines the cultural logic of uncertainty in China's children's fashion industry, and shows why familiar coordination devices, such as Order Fairs, remain peripheral to this market.
This study introduces hyperbolic embedding methods to map social networks and semantic combinations into comparable Poincaré disk spaces, where radius captures hierarchy and angle captures diversity. Analyzing 21st Century physics publications from the American Physical Society, the research demonstrates that dense, centralized collaboration among institutions is associated with a sharp reduction in the diversity of ideas explored.
Analyzing citation patterns from 5 million papers (1990–2024) and 1.1 million authors, this study reveals that scientist prestige and experience have opposite associations with AI knowledge use. Prestige correlates with citing highly visible, mainstream AI work — converging attention toward a narrow set of "hits" — while experience is associated with drawing on less visible, longer-tail AI literature. This divergence persists both within and outside Computer Science, suggesting that prestige may drive knowledge concentration while experience preserves the diversity essential for long-term scientific discovery.
Through mathematical decomposition and large-scale bibliometric evidence from 49 million journal articles, this study clarifies what the Disruption Index truly measures: a paper's ability to displace the dominant idea represented by its most-cited reference. Because citation counts among references follow Zipf's Law, competition is overwhelmingly shaped by the focal paper and its top predecessor, making the metric robust to "citation inflation." Breakthroughs, from this perspective, arise not only from generating novel ideas but from replacing established ones.
Based on an analysis of 41 million papers (1965-2024), this research challenges the prevailing recombinant growth theory. It uncovers a robust negative correlation between novelty and disruption: while "atypical combinations" of distant knowledge generate immediate attention, they consistently fail to displace entrenched ideas. True breakthroughs emerge not from broad search, but from focused, local displacement. This structural tension explains the modern "productivity-progress paradox," where science produces more novel combinations than ever, yet fundamental breakthroughs remain rare.
Complete academic and professional information