TurboLynx Dissertation Thesis

Jihoo Park · 2026-08-19T01:47:35+00:00

TurboLynx Dissertation Thesis available here: https://oasis.postech.ac.kr/handle/2014.oak/136963 TurboLynx is a graph analytics engine that combines schemaless property graphs with fast, general-purpose analytics by embedding schemaless support across storage, query processing, and optimization. It stores data as columnar graphlets—clusters of records with similar schemas—along with a graphlet-aware optimizer and processor to speeds up both traversal and analytical operations. In evaluations using LDBC SNB Interactive, TPC-H, and DBpedia workloads, it achieved up to 183.9x speedups over leading native graph databases and up to 41.27x faster than top relational analytics engines, while fully supporting the Cypher query language. The design demonstrates that high-performance analytics can coexist with flexible, evolving graph schemas.

From: Jihoo Park | Date: Wed, 19 Aug 2026 01:47:35 +0000 #
Subject: TurboLynx Dissertation Thesis
TurboLynx Dissertation Thesis available here: https://oasis.postech.ac.kr/handle/2014.oak/136963 TurboLynx is a graph analytics engine that combines schemaless property graphs with fast, general-purpose analytics by embedding schemaless support across storage, query processing, and optimization. It stores data as columnar graphlets—clusters of records with similar schemas—along with a graphlet-aware optimizer and processor to speeds up both traversal and analytical operations. In evaluations using LDBC SNB Interactive, TPC-H, and DBpedia workloads, it achieved up to 183.9x speedups over leading native graph databases and up to 41.27x faster than top relational analytics engines, while fully supporting the Cypher query language. The design demonstrates that high-performance analytics can coexist with flexible, evolving graph schemas.
[ Reply ]

[ Back to Group: comp.databases.turbolynx ]

Keyboard Shortcuts: [ j/n next | k/p prev | g top | r reply | u up ]