HPCSW 2011

Call for Papers

Over the last several years, there has been an increase of research in parallel semantic web data processing (SWDP) in the semantic web community as well as burgeoning interest in the high-performance computing (HPC) community. As specific examples, use of high-performance computing won the 2009 Billion Triple Challenge, and a parallel inference engine won the 2010 IEEE SCALE Challenge. The goal of the High-Performance Computing for the Semantic Web (HPCSW) workshop is to facilitate synergy between the HPC and semantic web communities as well as between academia and industry to further scalability of SWDP.

HPCSW is a full-day workshop that will begin with presentations of technical papers submitted for publication and will end with a discussion among (but not limited to) invited participants. Proceedings will be published online, and a selection of revised papers will be published in a joint Lecture Notes in Computer Science post-proceedings volume. Papers should present work in which HPC in some form (e.g., parallelism, supercomputers, FPGAs, etc.) is employed to improve SWDP (any kind of processing of semantic web data). Broad topics for papers include (but are not limited to):

• Parallelizing SWDP.
• Exploiting HPC architectures for SWDP.
• Employing parallel graph algorithms for SWDP.
• Benchmarks for SWDP from a HPC perspective.

The following questions will guide (but not limit) the discussion:

• How important is it to have high-performant applications for the semantic web?
• What are the boundaries of HPC for SWDP?
• What SWDP lends itself to HPC and/or parallel processing?
• What pre-existing work in HPC can be leveraged for SWDP?
• What are the tradeoffs between commodity HPC and specialized HPC for SWDP? (e.g., MapReduce vs. "hand-coded", commodity clusters vs. supercomputers)