E192-02 - Forschungsbereich Databases and Artificial Intelligence
-
Journal:
IEEE Access
-
ISSN:
2169-3536
-
Date (published):
2024
-
Number of Pages:
14
-
Publisher:
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
-
Peer reviewed:
Yes
-
Keywords:
Data models; data processing; data structures; database systems; databases; metadata; NoSQL databases; Graphs; raph data management
en
Abstract:
Graph serialization is a critical aspect of advancing graph-oriented systems and applications. Despite the importance of standardized serialization for property graphs, there is a lack of a universal format encompassing all essential features of graph database systems. This study introduces YARS-PG, a simple, extensible, and platform-independent serialization format tailored for property graphs. YARS-PG supports all the features permitted by current property graph-based database systems and is compatible with other graph-oriented databases and tools. We delineate the design requirements of YARS-PG by detailing both functional and non-functional aspects. Besides the basic features of property graph data, YARS-PG supports schema definition, metadata, metaproperties, variables, and graph definitions. Moreover, we discuss extensions of YARS-PG, demonstrating its flexibility through canonicalization techniques. Our comparative analyses with existing formats provide valuable insights, emphasizing the unique strengths that distinguish YARS-PG in the realm of graph data interchange. This paper serves as a definitive guide to YARS-PG, unraveling its complexities and showcasing its potential as a communication protocol, a data storage format, and a messaging specification.