Bio

Wu Lingfei is a PhD candidate in media and communication at the City University of Hong Kong. His research interests include complex network research and collective human behavior. He is now involved in a project exploring the dynamic evolution of online social networks.

He is especially interested in applying tools of statistical physics in analyzing social systems. “How does order emerge from random human behaviors?” is the core question he tries to answer. He think NKS summer school will not only offer him some powerful tools in exploring statistical laws in social systems, such as cellular automata, but will also bring him to a exciting place where he can meet peers who share the same academic interests.

Project: A Sample of Network Structure

In our society, people link to each other, and change the links from time to time. We can model it by using a time-variant social network model. Then the principle of computational equivalence will naturally lead us to an interesting question: what are the inputs and outputs if we regard network evolution as a computational process? One possible answer is that the “social network computer” takes attributes of its members as input, and outputs a link structure in which social members are classified into clusters. There is a saying that describes this process vividly: “birds of a feather flock together”.

In our project, we set a very simple local rule for nodes in the network: in every step of evolution, node x has probability p to link to node y who is two links away and has the minimum distance from x in the attribute-space, or x can randomly select y from all unconnected nodes with probability 1-p. We then investigate the network structure that this simple rule finally leads to. By comparing the network structure evolution process in simulation (which can be measured by several statistical indices such as cluster coefficient, degree distribution, etc.) to the network structure evolution in empirical data (collected by tracing the time-variant structure of an online social network community), we can test our assumption that social network evolution is a computational process.

Some problems indicate directions for further research. Is it possible to emulate network evolution with cellular automata or a Turing machine? Then the computational process might be uncovered more clearly. Is it possible to build an “open system” in simulation, which enables online players to join in the evolution game and make use of their data to promote the network evolution (just like Richard Dawkins’s program “The Blind Watchmaker”)? Such an open system may capture some very important characters of natural evolution.

There are many interesting phenomena calling for explanation in the way of understanding nature’s evolution as computation. I hope more and more people will gather together and cooperate with each other to work on such a topic. Networking is an efficient way to help us human beings survive in natural selection, and it will be an efficient way to help us understand ourselves and the universe.

Favorite Four-Color Totalistic Cellular Automaton

Rule 658204