1RDF Needs Better Database Database science is all one, laws of physicsapply to all RDF is relational in spirit, with so...
2Virtuoso 7 Brings the Best of anAnalytics DB to RDF Column store, 6 to 14 bytes a quad, dependingon data, 3x more compac...
3State Of Play Star Schema Benchmark Single Server, 30G Virtuoso SQL: 8.9s MonetDB SQL: 18.9s Virtuoso SPARQL: 45.4s...
4State Of Play (continued) Run with RDF in Virtuoso and run circles aroundmany RDBMS Use task-specific schema for perfor...
5Big data is not only queries Many things (e.g., graph analytics) cannot beexpressed as queries Data preparation and ETL...
6Future convergence RDF is, in the end, about relations The URI, the Superkey, is the ultimate silobreaker Get Schema-l...
6Future convergence RDF is, in the end, about relations The URI, the Superkey, is the ultimate silobreaker Get Schema-l...
Upcoming SlideShare
Loading in...5
×

ESWC 2013 Panel - Semantic Technologies for Big Data Analytics: Opportunities and Challenges

4,746

Published on

Big Data keeps climbing its hype-cycle hill, now above semantics and many other terms. But what do these in fact mean? In the leading books about Big Data, the word Semantics does not occur.

Published in: Technology, Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
4,746
On Slideshare
0
From Embeds
0
Number of Embeds
42
Actions
Shares
0
Downloads
0
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Transcript of "ESWC 2013 Panel - Semantic Technologies for Big Data Analytics: Opportunities and Challenges"

  1. 1. 1RDF Needs Better Database Database science is all one, laws of physicsapply to all RDF is relational in spirit, with some extracomplications To realize RDF’s potential, it must meet the bestin databasing, no separation of communities andno reinventing wheels© 2013 OpenLink Software, All rights reserved.
  2. 2. 2Virtuoso 7 Brings the Best of anAnalytics DB to RDF Column store, 6 to 14 bytes a quad, dependingon data, 3x more compact than before Vectored execution Get the best of the metal,take advantage of locality 3x the thread speedover previous Intra-query parallelization everywhere Scale out with flexible deployment, more parallel© 2013 OpenLink Software, All rights reserved.
  3. 3. 3State Of Play Star Schema Benchmark Single Server, 30G Virtuoso SQL: 8.9s MonetDB SQL: 18.9s Virtuoso SPARQL: 45.4s 2 Node Cluster, 300G Virtuoso SQL: 43.4s© 2013 OpenLink Software, All rights reserved.
  4. 4. 4State Of Play (continued) Run with RDF in Virtuoso and run circles aroundmany RDBMS Use task-specific schema for performance on thelevel of the best in SQL analytics Gap between SPARQL (schema-less) and SQL(schema-first) closing now 5x, going to 3x parity is reachable via self-tuning physical data layout© 2013 OpenLink Software, All rights reserved.
  5. 5. 5Big data is not only queries Many things (e.g., graph analytics) cannot beexpressed as queries Data preparation and ETL pipeline Map reduce and Bulk Synchronous Processing (BSP) frameworks Virtuoso can do all that with parallel SQLprocedures and Java hosting© 2013 OpenLink Software, All rights reserved.
  6. 6. 6Future convergence RDF is, in the end, about relations The URI, the Superkey, is the ultimate silobreaker Get Schema-less flexibility, self-describing data,and schema first performance, through adaptivephysical layout Open RDF data for use by the SQL tool chain© 2013 OpenLink Software, All rights reserved.
  7. 7. 6Future convergence RDF is, in the end, about relations The URI, the Superkey, is the ultimate silobreaker Get Schema-less flexibility, self-describing data,and schema first performance, through adaptivephysical layout Open RDF data for use by the SQL tool chain© 2013 OpenLink Software, All rights reserved.

×