Sign inSign up
Boosting Performance of Web Search Engines Using Query Logs thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Boosting Performance of Web Search Engines Using Query Logs

Published on Feb 25, 20075780 Views

Fabrizio Silvestri

Future of Web Search 2006 - Barcelona

Related categories

Information RetrievalWeb Mining

Chapter list

Boosting Performance of Web Search Engines using Query Logs00:00
Introduction00:29
What is a Query Log01:17
Our Logs01:49
What Does it Contain?02:59
What Does it Contain?03:38
How We Exploited04:34
Caching Policies05:12
Probability Driven Caching05:32
Static Dynamic Caching06:44
SDC Sections07:02
SDC and Prefetching07:39
Hit Ratio08:23
Throughput09:19
Collection Partitioning and Selection10:42
Innovations11:31
Contingency Matrix12:35
Co-clustering Results13:02
Experiments13:44
Precision15:10
Precision16:11
Better Term Partitioning17:06
Why Term Partitioning?18:10
Encouraging Figures...18:42
Discouraging Results19:02
Term Partitioning Enhancements19:47
Training and Test Sets20:21
Servers per Query20:57
List Replication21:59
Query per Servers - Replicated Lists22:10
Load Balancing (No Buffer Cache)22:40
Load Balancing - Network is Dominant22:42
Evaluating the Topic Shift Effect23:44
Various Training Times24:26
Thank You!24:29