Benchmarking Federated SPARQL Query Engines: Are Existing Testbeds Enough? thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Benchmarking Federated SPARQL Query Engines: Are Existing Testbeds Enough?

Published on Dec 03, 20122985 Views

Testbeds proposed so far to evaluate, compare, and eventually improve SPARQL query federation systems have still some limitations. Some variables and con gurations that may have an impact on the beha

Related categories

Chapter list

Benchmarking Federated SPARQL Query Engines: Are Existing Testbeds Enough?00:00
Motivation - 100:23
Motivation - 201:06
Motivation - 301:23
Our Challenges01:34
Independent Variables02:47
Dependent/Observed Variables03:15
Query Dimension - 103:35
Query Dimension - 203:53
Federated SPARLQ Queries Plan 1 (Exclusive Groups + Left Linear Plans)04:20
Plan 2 (Bushy Tree)05:22
Different federated plans05:34
Data Dimension06:07
Data Fragmentation06:17
Horizontal Fragmentation06:47
Vertical Fragmentation07:47
Data Fragmentation and Replication Effects08:14
Partitioning and Data Distribution LD10 (Perfect Network)09:26
Data Distribution11:07
Platform Dimension11:47
Endpoint Dimension12:18
Endpoint Dimension-Network Latency Gamma Distribution12:35
Endpoint Dimension13:08
Evaluation Hypothesis14:03
Experimental study - 114:28
Experimental study - 216:06
Experimental Evaluation16:58
Experimental Evaluation: Spearman’s Correlation Cross Domain17:26
Experimental Evaluation: Spearman’s Correlation Linked Data19:17
Experimental Evaluation: Spearman’s Correlation Life Science19:38
Experimental Evaluation: Spearman’s Correlation Complex Queries19:52
Conclusions21:03
Future Work21:49