Educations

• 1998  Seoul National University (Ph. D. in Physics)
• 1993  Seoul National University (M.S. in Physics)
• 1991  Seoul National University (B.S. in Physics)
   

• 2001~Present   Assistant Professor / Associate Professor / Professor, KAIST

• 2001~2001       Research Assistant Professor, Univ. of Notre Dame

• 1998~2000       Post. Doc., Univ. of Notre Dame
   

LAB

Complex Systems and Statistical Physics Laboratory
KAIST Physics Department’s Complex Systems and Statistical Physics Laboratory (CSSPL), led by Professor Hawoong Jeong, has produced significant work in four key areas: fractals, network science, big data, and artificial intelligence. The following sections summarize their research contributions for each keyword.

• Fractals
  Professor Jeong began his career studying fractals in statistical physics before expanding into network science. His doctoral work focused on fractal surface growth, and those insights laid the groundwork for later investigations into complex network structures. In CSSPL, researchers discovered that the physical layout of Internet topology forms a fractal set determined by global population density patterns. This work showed that network nodes exhibit fractal structures and that the placement of links arises from a competition between preferential attachment and linear distance dependency, rather than simple exponential decay. Additionally, studies of wave pattern formation in sputter erosion have elucidated the dynamic connections between fractal geometry and nonlinear phenomena.
• Network Science
  Professor Jeong is recognized as one of the founding figures of complex network science. Between 1999 and 2001, during his postdoctoral work at the University of Notre Dame with Professor Albert-László Barabási, he co-discovered “hubs” and power-law degree distributions in the World Wide Web. That landmark series of papers in Nature effectively launched the field of network science. In CSSPL’s subsequent work, researchers have analyzed network robustness under failures and attacks, mapped large-scale metabolic networks, and studied centrality and lethality in protein–protein interaction networks. Notably, they applied physics to prove the “Price of Anarchy” in road traffic networks—showing how individually rational driver choices create systemic congestion costs—a result highlighted both in Physical Review Letters and in The Economist.
• Big Data
  CSSPL integrates big-data science with complex network theory to drive innovative analyses. In one line of research, they quantified the complexity of paintings using information-theoretic measures—such as roughness indices, image entropy, and fractal dimension—to interpret art-historical evolution through data. Their social-network analyses have characterized topological features of large online platforms (Twitter, Facebook). They also developed methods using Google search results to build weighted networks of connections among parliamentarians, identifying hub figures in political networks. A novel approach combining text mining and network analysis uses search-engine co-occurrence weights to forecast emerging trends.
• Artificial Intelligence
  CSSPL’s AI research focuses on deep-learning approaches for understanding complex systems. They introduced AgentNet, a model-free, data-driven framework that uncovers hidden interactions in multi-agent systems. AgentNet employs graph-attention networks and per-variable attention to model interactions and has successfully captured dynamics in cellular automata, Vicsek flocking, and active Ornstein–Uhlenbeck particles. Experiments with real flocking bird data confirmed that AgentNet can reproduce collective behaviors qualitatively identical to nature. The lab also developed ConservNet, a neural network that automatically discovers conservation laws in grouped data. Most recently, they applied neural-network models to explain how intrinsic musical instincts can emerge in the human brain without explicit learning, a study published in Nature Communications.

Research Interest

• Complex Systems and Statistical Physics
  - Artificial Intelligence for Science
  - Data Science
  - Network Science
  - Computational methods in Statistical Physics
  - Dynamics of fluctuating interfaces and growing surfaces