Modeling Web Content Dynamics

Observing changing information sources

Your brain is good at this

Ingredients

A meaning for “up to date”

(a,b)-currency has meaning in many contexts

Slide 7

Collecting web page data

Sampling issues

Data in our collection

“Lifetimes” vs. “ages”

Determining dynamics from the time data

Sampling the lifetime distribution

Web page age CDF

Empirical lifetime distribution

When do changes happen?

Distribution of mean change times

Slide 18

(a,b)-currency for Poisson source

Probability a of b-currency over a collection

Index performance surface:
a as a function of T, n=b/T

Slide 22

Bandwidth needed for
(0.95, 1-week) currency

Empirical search engine (a,b)-currency

A calculus for (a,b)-currency

Summary

Mathematics of
“Semantic Hacking”

Problem

Distribution of information

What makes a good mystery/thriller?

Ongoing research

Acknowledgements