• Corpus ID: 32864328

On the design of a Globally Distributed, Locally Compressed Knowledge Base System

@inproceedings{Cur2017OnTD,
  title={On the design of a Globally Distributed, Locally Compressed Knowledge Base System},
  author={Olivier Cur{\'e}},
  booktitle={CIDR},
  year={2017}
}
Most big data processing are addressed by distributing the data and the execution of programs over a cluster of commodity hardware. These processing are generally handled by a cluster computing engine based on the MapReduce approach, e.g., Apache Hadoop. We consider that many big data related operations can be performed on a single machine if sufficient effort is put into compressing the data and ensuring that the programs can efficiently manipulate the compressed data, ideally in a…