=Paper=
{{Paper
|id=Vol-2369/short06
|storemode=property
|title=On Directly Mapping Relational Databases to Property Graphs
|pdfUrl=https://ceur-ws.org/Vol-2369/short06.pdf
|volume=Vol-2369
|authors=Radu Stoica,George Fletcher,Juan F. Sequeda
|dblpUrl=https://dblp.org/rec/conf/amw/StoicaFS19
}}
==On Directly Mapping Relational Databases to Property Graphs==
https://ceur-ws.org/Vol-2369/short06.pdf
On Directly Mapping Relational Databases to
Property Graphs
Radu Stoica1 , George Fletcher1 , and Juan F. Sequeda2
1
Eindhoven University of Technology, the Netherlands
{r.a.stoica@student., g.h.l.fletcher@}tue.nl
2
Capsenta, Austin, Texas, USA
juan@capsenta.com
Abstract. Much of the data found in practice resides in relational DBs.
However, many contemporary analytical tasks are performed on graphs.
Property graphs are currently one of the most prevalent data models
for graph data management in industry. Therefore, a key challenge is to
understand the fundamental relationships between relational databases
and property graph databases. This paper reports our ongoing work
towards understanding these relationships by proposing R2PG-DM, a
direct mapping of relational databases to property graphs. Given a re-
lational database schema and instance, a direct mapping generates a
corresponding property graph instance. The semantics of our mapping
is defined using Datalog. Our work is inspired by existing approaches
for direct mappings of relational databases into earlier graph data mod-
els. Future work is to study our mapping with respect to fundamental
properties such as information and query preservation.
1 Introduction
A major class of contemporary data analytics focuses on gaining insights from the
rich complex patterns found in graph-structured data collections such as social,
communication, financial, biological, and mobility networks [1]. For example,
investigative journalists have recently found, through graph analytics, surprising
social relationships between executives of companies within the Offshore Leaks
financial social network data set, linking company officers and their companies
registered in the Bahamas.3 Property graphs (PG) are currently one of the most
prevalent data models for the management of such data in industry [1]. A PG
is an edge- and node-labeled directed multigraph where both edges and nodes
have associated sets of properties, i.e., key-value pairs.
The Offshore Leaks PG was constructed as a mapping from relational database
(RDB) sources.4 Indeed, much of the data found in practice resides in RDBs.
Therefore, a key challenge is to understand the fundamental relationships be-
tween RDBs and PGs. A crucial first step in exploratory graph analytics on
3
International Consortium of Investigative Journalists. https://offshoreleaks.icij.org/
4
https://www.icij.org/blog/2013/06/how-we-built-offshore-leaks-database/
2 R. Stoica et al.
RDBs is to transform the database into a PG. Only then can the basic graph
structure be explored and new relationships discovered using contemporary graph
database systems.
This motivates the study of direct mappings from RDBs to PGs. A direct
mapping is a transformation from RDBs to PGs which (1) is domain and schema-
independent, i.e., works regardless of the source database schema and instance,
and (2) transforms the content of the source instance into a target instance, i.e.,
given a RDB instance generates a corresponding PG instance.
State of the art. Most approaches to graph analytics over relational data
extract graphs as user-specified views on relational data, e.g., [3,8]. This requires,
however, that the user already fully understands the desired graph view, which
is overly restrictive in the common case of exploratory graph analytics.
The study of direct mappings from relational to property graphs is not as
well-developed. Of the small handful of approaches, the focus has been on opti-
mized layout for efficient query processing or mappings which are lossy or obfus-
cate the input RDB schema [4,5,7,9]. Furthermore, all current direct mappings
are defined procedurally (i.e., defined with pseudo-code), making it difficult to
formally reason about their correctness and other basic properties.
Our contributions. In this short note, we report on our work-in-progress on
R2PG-DM, a declarative direct mapping from RDB to PG. R2PG-DM losslessly
transforms both the source instance and schema while also intuitively preserving
the original structure of the input RDB. Our work is inspired by the approach of
Sequeda et al. to direct mappings from relational to RDF graphs, the W3C stan-
dard for sharing graph-structured data on the web [6]. We present R2PG-DM by
example and outline the basic research questions we are currently investigating
in our study of the connections between RDB and PG.
2 RDBs, PGs, and direct mappings
Let D be an enumerable set of data values containing the special value null, and
let A be an enumerable set of attribute names containing the special attribute
tid. An RDB schema is a triple S = (R, att, Σ), where R is a finite set of relation
names, att is a function assigning to each r ∈ R a finite set att(r) ⊆ A \ {tid},
and Σ is a finite set of primary and foreign keys over R and att.5 For r ∈ R, an
r-tuple is a function t : att(r) ∪ {tid} → D such that t(tid) 6= null. An instance I
of S is an assignment to each S r ∈ R of a finite set I(r) of r-tuples satisfying Σ,
such that for distinct t, t0 ∈ r∈R I(r) it holds that t(tid) 6= t0 (tid). An example
schema and instance is given in Figure 1 (top left).
A property graph is a structure (V, E, e, `, p) where V is a finite set of vertices,
E is a finite set of edges, e is a function assigning an ordered pair of vertices
to each edge (i.e., the source and target vertices of the edge, resp.), ` assigns a
finite set of labels to each vertex and edge (from some domain of labels), and
p assigns a finite set of key-value pairs to each vertex and edge. An example
property graph is given in Figure 1 (bottom left).
5
https://en.wikipedia.org/wiki/Foreign_key
On Directly Mapping Relational Databases to Property Graphs 3
Fig. 1. (top left) An RDB (S, I), where primary key attributes are underlined, partnerA
and partnerB of Partners are foreign keys to name of Person, and location of Partners is
a foreign key to cname of City. (bottom left) The property graph R2PG-DM(S, I) and
(right) its representation with predicates V ertex, Edge, and P roperty.
Let PG denote the set of all PGs and RDB denote the set of all pairs (S, I)
where S is an RDB schema and I is an instance of S. A direct mapping is a total
function from RDB to PG.
3 The R2PG-DM direct mapping
We next informally present R2PG-DM, a direct mapping which addresses the
shortcomings of current solutions discussed in Section 1. With R2PG-DM, RDB
relations are interpreted as classes of “things” (i.e., labeled vertices) and foreign-
key relationships are interpreted as edges between these things. In particular:
– each input r-tuple t is represented in the output graph by a vertex v (iden-
tified by t(tid)), which is labeled with r, the name of the relation to which t
belongs; furthermore, v has key-value pair (a, t(a)), for each a ∈ att(r).
– for each foreign key relationship f ∈ Σ (from, say, relation r to relation
s), for each pair of tuples t ∈ I(r) and t0 ∈ I(s) participating in f , this
relationship is represented by an edge with label r-s from vertex vt to vertex
vt0 , representing t and t0 , respectively.
4 R. Stoica et al.
Similarly, as there currently does not exist a standard PG schema language,
the input schema is also fully represented in the output PG, e.g., each relation
and each attribute of a relation is represented by a vertex, and foreign keys are
represented by edges between the relevant attributes of relations, etc. Figure 1
illustrates an RDB database (S, I) and PG translation R2PG-DM(S, I). We omit
here the translation of S to PG, and only illustrate the translation of I.
Clearly, each component of the R2PG-DM mapping can be specified by a
declarative non-recursive Datalog query [2] over the source DB represented using
a fixed set of predicates (see Section 4.1 “Storing relational databases” of Sequeda
et al. [6]), with the target PG represented by three predicates V ertex, Edge, and
P roperty capturing the vertices, edges, and the key-value properties associated
with vertices and edges, respectively. This is illustrated in Figure 1 (right).
4 Ongoing research
Our ongoing work proceeds along three lines. First, we are proceeding with a
formal study of R2PG-DM, establishing basic properties such as information and
query preservation [6]; we hypothesize that R2PG-DM generates a graph that is
isomorphic as a graph generated by the direct mapping of [6]. Moreover, we are
studying how to extend R2PG-DM in order to generate a graph with properties
on edges, taking full advantage of the data model. Second, we are developing
practical tools for efficient and scalable R2PG-DM, to be made available as
open source code. Third, we aim to extend the R2PG-DM approach to support
customized mappings for the cases where there is a schema defined on the target
instance (e.g., mapping the relational schema S to an equivalent PG schema).
This builds upon ongoing efforts towards standards for PG schema languages.6
References
1. Angela Bonifati, George Fletcher, Hannes Voigt, Nikolay Yakovets. Querying
Graphs. Morgan & Claypool, 2018.
2. Todd J. Green et al. Datalog and recursive query processing. Foundations and
Trends in Databases 5(2):105-195, 2013.
3. Mohamed S. Hassan et al. Extending in-memory relational database engines with
native graph support. In EDBT 2018, pages 25-36.
4. Ognjen Orel, Slaven Zakošek, Mirta Baranović. Property oriented relational-to-
graph database conversion. Automatika 57(3): 836-845, 2016.
5. Subhesh Pradhan, Sharma Chakravarthy, Aditya Telang. Modeling relational data
as graphs for mining. In COMAD 2009.
6. Juan F. Sequeda, Marcelo Arenas, Daniel P. Miranker. On directly mapping rela-
tional databases to RDF and OWL. In WWW 2012, pages 649-658.
7. Roberto De Virgilio, Antonio Maccioni, Riccardo Torlone. R2G: a tool for migrating
relations to graphs. In EDBT 2014, pages 640-643.
8. Konstantinos Xirogiannopoulos, Amol Deshpande. Extracting and analyzing hidden
graphs from relational databases. In SIGMOD 2017, pages 897-912.
9. Kang Min Yoo, Sungchan Park, Sang-Goo Lee. RDB2Graph: a generic framework
for modeling relational databases as graphs. In JIST 2014, pages 148-151.
6
https://www.w3.org/Data/events/data-ws-2019/