Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 1907.07405

In-Depth Benchmarking of Graph Database Systems with the Linked Data Benchmark Council (LDBC) Social Network Benchmark (SNB)

Florin Rusu·Zhiyi Huang
View original source ↗Read source document ↗Export metadata ↓

Abstract

In this study, we present the first results of a complete implementation of the LDBC SNB benchmark -- interactive short, interactive complex, and business intelligence -- in two native graph database systems---Neo4j and TigerGraph. In addition to thoroughly evaluating the performance of all of the 46 queries in the benchmark on four scale factors -- SF-1, SF-10, SF-100, and SF-1000 -- and three computing architectures -- on premise and in the cloud -- we also measure the bulk loading time and storage size. Our results show that TigerGraph is consistently outperforming Neo4j on the majority of the queries---by two or more orders of magnitude (100X factor) on certain interactive complex and business intelligence queries. The gap increases with the size of the data since only TigerGraph is able to scale to SF-1000---Neo4j finishes only 12 of the 25 business intelligence queries in reasonable time. Nonetheless, Neo4j is generally faster at bulk loading graph data up to SF-100. A key to our study is the active involvement of the vendors in the tuning of their platforms. In order to encourage reproducibility, we make all the code, scripts, and configuration parameters publicly available online.

Explore related subjects

cs.DB

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Florin Rusu, Zhiyi Huang. 2019-07-17. In-Depth Benchmarking of Graph Database Systems with the Linked Data Benchmark Council (LDBC) Social Network Benchmark (SNB). https://arxiv.org/abs/1907.07405

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Access Paths for Efficient Ordering with Large Language Models

In this work, we present the \texttt{LLM ORDER BY} semantic operator as a logical abstraction and conduct a systematic study of its physical implementations. First, we propose several improvements to existing semantic sorting algorithms and introduce a semantic-aware external merge sort algorithm. Our extensive evaluation reveals that no single implementation offers universal optimality on all datasets. From our evaluations, we observe a general scaling relationship between sorting cost and the ordering quality for comparison-based algorithms. Building on these insights, we design a budget-aware optimizer that utilizes heuristic rules, LLM-as-Judge evaluation, and consensus aggregation to dynamically select the near-optimal access path for LLM ORDER BY. In our extensive evaluations, our optimizer consistently achieves ranking accuracy on par with or superior to the best static methods across all benchmarks. We believe that this work provides foundational insights into the principled optimization of semantic operators essential for building robust, large-scale LLM-powered analytic systems.

cs.DB↗
arXiv2026

Expressive Power of Property Graph Constraint Languages

We present the first principled and systematic study of the expressive power of property graph constraint languages, focused on the recent PG-Keys language, set to inform the upcoming revision of the GQL standard. To this end, we position PG-Keys within the broader landscape of existing formalisms. In particular, we compare PG-Keys with two core property graph constraint languages: Graph Functional Dependencies (GFD) and Graph Generating Dependencies (GGD). One hurdle is that these formalisms allow different kinds of graph pattern languages and data predicates. To make a fair comparison, based on their structural differences only, we first present a unifying framework. Within this framework, we consider conjunctive regular path queries (CRPQ) as graph patterns with equality and inequality predicates. We then identify well-behaved fragments, establish expressiveness inclusion, and prove separation results, yielding a complete and strict hierarchy of expressive power. The results identify precisely when PG-Keys provide strictly greater expressive power, clarifying their place among state-of-the-art property graph constraint formalisms.

cs.DB↗
arXiv2026

Answering Conjunctive Queries with Aggregations under Updates

Dynamic query processing keeps query answers up to date during insertions and deletions. For conjunctive queries (CQs) under set semantics, the classes maintainable in constant amortized time are known exactly: the $q$-hierarchical CQs under arbitrary updates, and the free-connex CQs under insertion-only updates. Many analytics tasks, including \textsf{SUM}/\textsf{COUNT} aggregations, provenance, and access control, are captured by evaluating a CQ over a positive commutative semiring. We thus ask whether aggregation changes what can be maintained efficiently, and if so, when. Under \emph{insertion-only} updates, it does: the boundary retreats from free-connex to a new class we call \emph{strong-connex}, with $q\text{-hierarchical} \subsetneq \text{strong-connex} \subsetneq \text{free-connex} \subsetneq \text{acyclic}$. For every \emph{strictly monotone} semiring, including the sum-product and tropical semirings, no free-connex but non-strong-connex CQ is maintainable in $O(|D|^{1/2-ε})$ time under the OuMv and OMv conjectures, whereas every strong-connex CQ is maintainable in $O(1)$ amortized time over every semiring. Under \emph{arbitrary} updates, the boundary stays at the $q$-hierarchical CQs for every semiring with $O(1)$-deletable aggregates, and maintenance over any semiring is at least as hard as over the Boolean semiring. We further strengthen the lower bounds to semirings that fall outside the class and to query with different \emph{height} and \emph{dimension}, under the combinatorial $k$-clique and generalized OuMv conjectures. All upper bounds come from a single framework, obtained by adapting CROWN to annotated relations; together with the lower bounds, they yield dichotomies parameterized by both the query and the semiring, recovering the Boolean results as a special case.

cs.DB↗