0.25
0.5
0.75
1.25
1.5
1.75
2
Querying Datasets on the Web with High Availability
Published on Dec 19, 201411579 Views
As the Web of Data is growing at an ever increasing speed, the lack of reliable query solutions for live public data becomes apparent. sparql implementations have matured and deliver impressive perfor
Related categories
Chapter list
Querying datasets on the Web with high availability00:00
Somebody comes to you with a high-quality Linked Open Data set:00:03
Untitled00:11
Querying Linked Data on the Web always involves trade-offs.00:52
Querying Linked Data live on the Web becomes affordable00:56
We formalized and evaluated client-side SPARQL querying through a triple-pattern interface.01:02
Querying datasets on the Web with high availability01:22
Querying datasets on the Web with high availability01:43
Untitled01:45
Linked Data Fragments are a uniform view on Linked Data interfaces.01:50
Each type of Linked Data Fragment is defined by three characteristics.03:17
Each type of Linked Data Fragment is defined by three characteristics.03:54
We designed a new trade-off mix with low cost and high availability03:57
A triple pattern fragments interface is low-cost and enables clients to query.04:04
A triple pattern fragments interface is low-cost and enables clients to query.04:25
DBpedia04:59
Triple patterns are not the final answer. No interface ever will be.05:24
Querying datasets on the Web with high availability05:44
Triple pattern fragment servers enable clients to be intelligent.05:55
Triple pattern fragment servers enable clients to be intelligent.06:07
Triple pattern fragment servers enable clients to be intelligent.06:46
Triple pattern fragment servers enable clients to be intelligent.06:51
How can intelligent clients solve SPARQL queries over fragments?06:58
Let’s follow the execution of an example SPARQL query.07:23
The client looks inside the fragment to see how to access the dataset.07:46
The client splits the query into the available fragments.07:56
The client gets the fragments and inspects their metadata.08:07
The metadata enables the client to choose the right starting point.08:42
Clients execute the query in 3 seconds on a highly available, low-cost server.09:32
We evaluated triple pattern fragments for server cost and availability.09:33
We evaluated triple pattern fragments for server cost and availability.09:57
The query throughput is lower, but resilient to high client numbers.10:13
The server traffic is higher, but requests are significantly lighter.10:35
Caching is significantly more effective, as clients reuse fragments for queries.11:10
The server uses much less CPU, allowing for higher availability.11:23
Servers enable clients to be intelligent, so they remain simple and light-weight.11:47
Querying datasets on the Web with high availability11:58
Triple pattern fragments allow querying live data with high availability.12:10
Like each Linked Data query method, triple pattern fragments have trade-offs.12:18
Experience the trade-offs yourself on the official DBpedia interfaces.12:38
Triple pattern fragments are easy: all software is available as open source.13:10
Querying is a client–server dialogue. Start exploring the entire spectrum.13:21
Thank you!14:08