The World Wide Web: An (Electrical) Engineering Tour de Force

Professor Vwani Roychowdhury
UCLA Electrical Engineering

Wednesday, April 9, 2008 at 1:00PM

54-134 Engineering IV Building
Refreshments Served

Abstract: The World Wide Web (WWW) has emerged as a symbiotic socioeconomic entity, enabling new forms of commerce and social intercourse, while being constantly updated and modified by the activities that it itself enables. A ubiquitous communication network with an ever increasing reach forms the backbone of the web and is the primary enabler. Without the advances made in diverse electrical engineering disciplines, including devices, communications, signal processing, and information theory, the web as we know it would not exist. The talk will provide a very brief introduction to the world before the web and some of the advances and visions that led to the emergence of the web. The focus of the talk, however, will not be centered on such no-brainer and post-mortem foundational links between electrical engineering and the web. Instead, we will focus on recent areas of research that involve the dynamics of, and information propagation on, the web, and how the information processing tools that we as Electrical Engineers learn and use can be used to both understand and influence the evolution and impact of the web.

For example, the Web is typically modeled as an evolving network whose nodes are web-pages and whose edges are directed URL links. Some of the questions we ask are: what are the macroscopic structures in this web graph? Are there realistic microscopic models of web growth that would give rise to some of the structures that are observed? Why would the structure matter to any of the users of the web, who just "Google it?" It turns out that one can indeed answer many of these questions using ideas from statistical mechanics, stochastic processes, random graphs, and combinatorics, all of which are fields that we as Electrical Engineers contribute to and are major practitioners of. Similarly, we will investigate the mechanisms of information propagation on the web: How does information spread through the web? What is the special nature (if any) of information driven traffic on the web and its impact on the physical layer backbone? What makes one web page popular and a similar page not so popular? What is the science behind the monetization secrets of the online giants such as Google, Yahoo! and MSN? It turns out that one can begin to answer many of these questions using principles of information dynamics, branching processes, game theory, and optimization. Of course, nothing is a one-way street, and we already find ideas and analytical tools developed in this context being fed back into answering questions in the parent fields.

The talk will end with a tantalizing peek at a fascinating question (without any answers of course): Can an alien species reverse engineer the content and structure of the web (undoubtedly, the largest dynamic repository of human activity logs) and create an oracle that would pass the Turing test? You get all this and more in about an hour's worth of your time. So bring your friends and your i-Pod.

Biography