tsior2022, 1907 ariner Repent Gartner. Licensed for Distribution Market Guide for Graph Database Management Solutions Published 24 May 2021 - ID G00737853 - 33 min read By Merv Adrian, Afraz Jaffri, and 1 more Data and analytics leaders are challenged with a multitude of existing and new DBMS choices for use with graph analytics use cases. This Market Guide describes available graph database management solutions Over Ww Key Findings = Graph database management solutions (OBMSs) are growing in awareness and popularity as they and their associated tools become more mature. However, itis often difficult for data and analytics professionals to distinguish between different implementation models, and to fit them to their use case. = Native and multimodel graph databases have the ability to scale to meet enterprise production requirements, including stability, availability and security, but native graph DBMSs offer performance gains when dealing with queries over very large graphs (typically billions of nodes). = The traditional split between RDF and property graph support is becoming less important for choosing a graph DBMS compared to features that support enterprise readiness, = Vendors in the graph DBMS market are expanding their stack to become platform plays for enterprise knowledge graphs or graph artificial intelligence (Al) with an associated ecosystem of products. 1 Managed as-a-service offerings from established and new vendors and cloud service providers significantly lower the barrier to entry for organizations that do not have graph expertise. Recommendations Data and analytics leaders responsible for data management solutions, and implementing and evolving operational and analytical infrastructure, should: hitpssiwwu.gartnerconvdectreprinis?__helo=108400980, 164ba5 !2a9fa a0 Tbdb62521916347 1642290390625.1642280909625.1642200309025.1... 11221510172022, 1907 Garner Reprint = Invest time to fully develop a graph use case before choosing a graph DBMS by defining business value, sample queries and transactional versus analytical processing requirements. = Assess the enterprise requirements for graph DBMS product functionality that go beyond prototypes by comparing security models, API support, ease of configuration and optimization, and end-user tools for visualization, query and schema building, = Distinguish between RDF or property graph models only where there is a clear need by evaluating data exchange, ontology reuse and reasoning requirements, = Test the ability of existing multimodel offerings to deliver the required analytics, scalability and performance before adding another vendor and product combination. = Deploy managed services where available to minimize the overhead of managing and maintaining graph DBMS, which may be unfamiliar to existing team members. Strategic Planning Assumption By 2025, graph technologies will be used in 80% of data and analytics innovations, up from 10% in 2021, facilitating rapid decision making across the enterprise. Market Definition Graph DBMSs provide support for graph data models, in native and virtual formats, including data loading, conversion, consistency and security, with the ability to scale up and scale out. They support multiple types of interactions ranging from simple node and edge traversal and triple pattern matching for transactional uses to complex multihop queries, reasoning and inference, and algorithms that run across the entire graph structure for analytical workloads. Market Description As in the overall DBMS market, graph DBMS use cases can be classified as operational or analytics (and increasingly involve both). Graph DBMS vendors have accordingly provided two types of product, Transactional graph DBMSs are optimized for retrieving small parts of one or multiple graphs, frequently through querying or traversal (transactional processing). Analytical graph DBMSs are optimized to perform operations on each node and edge in a graph, and aggregate results across the entire or large part of the graph. Some use-case examples are presented in Note 3. Another type of division that exists in the Graph DBMS market is in the way a graph is structured and how it can be queried: = Aproperty graph (also known as a labeled property graph) allows pairs of tuples (properties) to be added to any node or edge in a graph. Nodes can be grouped together using labels. hitpsiwwu.garinercondectreprinis?__helo=108400980, 164ba5'!2a9fa a0 Tbdb62521910347 1642290390625.1642280909625.1642200309025.1... 2:221510172022, 1907 ‘corer Rept = An RDF graph explicitly models every node and edge combination as a set of triples of the form