Awesome

Query-Centric Semantic Partitioning (SPARTI)

• Adaptively partition based on query-workload
• Precompute Bloom join between the most frequent triples joins (MF-TJ) combinations
• Partition related properties based on a greedy algorithm and a cost model
• Current version is implemented to run over <b>Apache Spark</b>