Leveraging Neo4j for Data Science: Evaluating Traversal Efficiency in GDS and APOC for Directed Acyclic Graphs

Jamil Saquer Hazim Shatnawi

18 April 2024

This study introduces an innovative approach to encoding and analyzing feature models within the Network Exploration and Optimization for Java (Neo4j) graph database, significantly enhancing the management of complex Software Product Lines (SPLs). We present a comparative analysis of traditional loading techniques against Neo4j's batch importer and the Awesome Procedures on Cypher (APOC) library, demonstrating the superior efficiency and effectiveness of our proposed methods, especially in handling large datasets. Our methodology extends beyond mere encoding; it capitalizes on Neo4j's Graph Data Science (GDS) library, employing Depth-First Search (DFS) and other advanced traversal techniques to navigate and manipulate these complex structures. The findings reveal not only a significant enhancement in the processing and analysis of feature models but also underscore the potential for more sophisticated SPL management strategies. By integrating innovative loading techniques, encoding strategies, and GDS traversal methods, this study lays a robust foundation for future advancements in the field.

Xiang Chen J. Atlee

1 Oktober 2023

Comprehensive analysis of a software product line (SPL) is expensive because the number of products to be analyzed is exponential in the number of the SPL’s features. To compensate, we analyze a model of the SPL rather than the source code, thereby reducing the size of the artifact under analysis. In this paper, we facilitate SPL analysis by lifting the Neo4j query engine to apply to an SPL model, so that a Neo4j query returns variability-aware results that cover all the SPL’s products. We used the lifted Neo4j to analyze five nontrivial SPLs (with respect to dataflows, control-flows, component interactions, etc.) and found that the overhead for returning variability-aware results for the full SPL, versus the results for just one product, ranges from 1.88% to 456%. In comparison to related work V-Soufflé (a lifted Datalog engine), lifted Neo4j is able to report complete path results whereas V-Soufflé reports only endpoints of paths. When both analyzers report the same results (e.g., endpoints of paths), lifted Neo4j is usually more efficient.

Oluwatosin Agbaakin Sydney Anuyah Victor Bolade

15 November 2024

This tutorial serves as a comprehensive guide for understanding graph databases, focusing on the fundamentals of graph theory while showcasing practical applications across various fields. It starts by introducing foundational concepts and delves into the structure of graphs through nodes and edges, covering different types such as undirected, directed, weighted, and unweighted graphs. Key graph properties, terminologies, and essential algorithms for network analysis are outlined, including Dijkstras shortest path algorithm and methods for calculating node centrality and graph connectivity. The tutorial highlights the advantages of graph databases over traditional relational databases, particularly in efficiently managing complex, interconnected data. It examines leading graph database systems such as Neo4j, Amazon Neptune, and ArangoDB, emphasizing their unique features for handling large datasets. Practical instructions on graph operations using NetworkX and Neo4j are provided, covering node and edge creation, attribute assignment, and advanced queries with Cypher. Additionally, the tutorial explores common graph visualization techniques using tools like Plotly and Neo4j Bloom, which enhance the interpretation and usability of graph data. It also delves into community detection algorithms, including the Louvain method, which facilitates clustering in large networks. Finally, the paper concludes with recommendations for researchers interested in exploring the vast potential of graph technologies.

Michael J. Carey Glenn Galvizo

13 Mei 2024

The increasing prevalence of large graph data has produced a variety of research and applications tailored toward graph data management. Users aiming to perform graph analytics will typically start by importing existing data into a separate graph-purposed storage engine. The cost of maintaining a separate system (e.g., the data copy, the associated queries, etc …) just for graph analytics may be prohibitive for users with Big Data. In this paper, we introduce Graphix and show how it enables property graph views of existing document data in AsterixDB, a Big Data management system boasting a partitioned-parallel query execution engine. We explain a) the graph view user model of Graphix, b) $\text{gSQL}^{++}$, a novel query language extension for synergistic document-based navigational pattern matching, and c) how edge hops are evaluated in a parallel fashion. We then compare queries authored in $\text{gSQL}^{++}$ against versions in other leading query languages. Finally, we evaluate our approach against a leading native graph database, Neo4j, and show that Graphix is appropriate for operational and analytical workloads, especially at scale.

Vladimir Kutuev Vlada Pogozhelskaya S. Grigorev + 1 lainnya

19 Desember 2023

We propose GLL-based context-free path querying algorithm which handles queries in Extended Backus-Naur Form (EBNF) using Recursive State Machines (RSM). Utilization of EBNF allows one to combine traditional regular expressions and mutually recursive patterns in constraints natively. The proposed algorithm solves both the reachability-only and the all-paths problems for the all-pairs and the multiple sources cases. The evaluation on realworld graphs demonstrates that utilization of RSMs increases performance of query evaluation. Being implemented as a stored procedure for Neo4j, our solution demonstrates better performance than a similar solution for RedisGraph. Performance of our solution of regular path queries is comparable with performance of native Neo4j solution, and in some cases our solution requires significantly less memory.