[论文解读] Semantic Property Graph for Scalable Knowledge Graph Analytics

Sumit Purohit, Nhuy Van+1 more|arXiv (Cornell University)|Sep 16, 2020

Semantic Web and Ontologies参考文献 15被引用 5

一句话总结

本文提出语义属性图（SPG），一种混合知识图谱模型，将重言式RDF图投影为带标签的属性图（LPG）格式，同时保留RDF本体以支持语义推理。通过利用高效的LPG存储与遍历机制，SPG减少了图的大小，提升了分析任务（如子图匹配）的可扩展性，相较于原生RDF表示实现了显著的压缩与性能提升。

ABSTRACT

Graphs are a natural and fundamental representation of describing the activities, relationships, and evolution of various complex systems. Many domains such as communication, citation, procurement, biology, social media, and transportation can be modeled as a set of entities and their relationships. Resource Description Framework (RDF) and Labeled Property Graph (LPG) are two of the most used data models to encode information in a graph. Both models are similar in terms of using basic graph elements such as nodes and edges but differ in terms of modeling approach, expressibility, serialization, and target applications. RDF is a flexible data exchange model for expressing information about entities but it tends to a have high memory footprint and inefficient storage, which does not make it a natural choice to perform scalable graph analytics. In contrast, LPG has gained traction as a reliable model in performing scalable graph analytic tasks such as sub-graph matching, network alignment, and real-time knowledge graph query. It provides efficient storage, fast traversal, and flexibility to model various real-world domains. At the same time, the LPGs lack the support of a formal knowledge representation such as an ontology to provide automated knowledge inference. We propose Semantic Property Graph (SPG) as a logical projection of reified RDF into LPG model. SPG continues to use RDF ontology to define type hierarchy of the projected graph and validate it against a given ontology. We present a framework to convert reified RDF graphs into SPG using two different computing environments. We also present cloud-based graph migration capabilities using Amazon Web Services.

研究动机与目标

• 为解决RDF图在大规模图分析中因内存占用过高和结构臃肿而导致的效率低下问题。
• 弥合RDF在语义表达能力与LPG在性能优势之间的差距，以支持可扩展分析。
• 通过统一框架实现高效、云规模的知识图谱生成与分析，同时保持本体语义。
• 在真实分析工作负载中展示SPG的实际优势，例如在对抗性活动检测中的子图匹配。

提出的方法

• 通过逻辑重言式将重言式RDF图转换为带标签的属性图（LPG）模型，同时保留本体语义。
• 采用包含三个核心组件的框架：查询解析器、序列化器和验证器，将SPARQL查询结果转换为SPG格式。
• 以GDF格式（含nodedef和edgedef部分）序列化SPG，实现高效存储与互操作性，未来将支持JSON格式。
• 使用专用的SPG验证器，依据RDF本体验证节点与边的类型，确保语义一致性。
• 将SPG生成集成至NLP流水线中，实现从新闻、社交媒体及事务性数据源获取真实世界数据。
• 利用Amazon Neptune支持云规模的SPG生成，并兼容分布式计算环境。

实验结果

研究问题

• RQ1 重言式RDF图能否在不损失语义表达能力的前提下，被高效地转换为属性图模型？
• RQ2 与原生RDF相比，SPG在图分析任务中的存储效率与查询性能如何？
• RQ3 SPG在子图匹配与网络对齐任务中，其可扩展性提升程度如何？
• RQ4 SPG能否在保持本体验证语义保真度的同时，实现更快的图遍历与更小的内存占用？

主要发现

• SPG相较于原生RDF显著减少了图的大小与序列化文件大小，尤其通过消除冗余元数据节点（如'Statement'和'rdf:type'）实现。
• SPG图的度分布比RDF图更均衡，表明其结构更具平衡性与可分析性。
• SPG中的子图查询所需边数更少——例如，SPG中的两跳楔形查询对比RDF中20条边、三跳的子图查询，体现出更强的结构紧凑性。
• SPG框架实现了高效、可扩展的图分析，具备更快的图遍历速度与更低的内存开销，尤其适用于实时与大规模工作负载。
• 将SPG集成至NLP流水线后，成功从包括新闻、社交媒体及事务性数据在内的多样化数据源生成紧凑且可分析的图谱。