NO.138 Graph Database Systems: Bridging Theory, Practice, and Engineering

Shonan Village Center

July 30 - August 2, 2018 (Check-in: July 29, 2018 )


  • Oskar van Rest
    • Oracle, USA
  • George Fletcher
    • TU Eindhoven, The Netherlands
  • Wook-Shin Han
    • Pohang University of Science and Technology, South Korea


Context and Motivation

Recent years have seen great advances in the study of data management solutions for massive graph-structured data sets. This has been stimulated by the increasing availability of large graphs in a broad variety of application domains such as social networks, biological networks, linked open data, communications networks, and mobility networks. Consequently, there has been a marked rise in demand for scalable solutions for the principled management of graph data. Rapid progress has been made on our understanding of: the theoretical foundations of fundamental topics such as graph query languages, graph analytics, and graph modelling; the engineering foundations of efficient and scalable graph intensive systems; and, the practical application and engineering of graph data management solutions in industry.

Through these advancements the graph database research community has now reached a first stage of maturity. However, this understanding and acquired wisdom is distributed across various disparate subcommunities in the field. The time is right for a community “checkpoint”, to share these experiences and insights across the rich and diverse areas of investigation in graph database systems. Indeed, a major outcome of this necessary checkpoint will be to consolidate our broad community understanding of the “first generation” of practical graph data management systems.

A second outcome of this taking stock and intense sharing of perspectives is to identify the major challenges and limiting factors in the realization of the next generation of graph database systems. Examples of such open challenges include: identifying appropriate graph schema formalisms and developing deeper our understanding of graph constraints (both in theory and practice); the efficient processing of recursive graph queries, such as the Regular Path Queries; developing practical syntaxes and engineering solutions for graph query languages supporting reasoning over data, e.g., in the property-graph model; practical human-in-the loop graph analytics and visual query methodologies; and, ensuring that we are asking the right questions driven by application domains and practical graph analysis.

Goals of the Meeting

The goal of this meeting is to take stock of the current state of the art in graph data management systems and to identify major open research challenges and directions, towards setting a community research agenda for the coming years. We place particular focus on building bridges between advances in the theory, engineering, and practical deployment of graph databases. For this broad discussion, we aim to bring together relevant leading researchers from both academia and industry, across these diverse subcommunities.

Concrete Outcomes Targeted

In addition to an in-depth NII technical report on the discussions and results of the seminar, we aim to write a community white paper for peer-reviewed publication, indicating the most important open challenges in the field identified during the meeting. Having an open bottom-up shared vision for the field will stimulate research and industry progress, setting the research agenda for the graph data management community for the coming years. The workshop will also lead to concrete action plans for international collaborations in research and longer-term international projects of broad ambition, including identifying concrete calls for proposals for research funding, to facilitate international collaboration.