Next: Curve-Fitting vs. Bound-Based Policy
Up: Recrawl Scheduling Based on
Previous: Bounding Risk
Online Experiments
We evaluate our online page revisitation policies empirically, using simulation over our web data sets described in Section 3.1.
There are three sets of experiments:
- Comparison between our two policies. (Section 5.1)
- Measurement of the space footprint. (Section 5.2)
- Comparison with previous work. (Section 5.3)
Figure 8:
Warm-up performance of online revisitation policies, on high-quality data set.
|
Subsections
Chris Olston
2008-02-15