Validate GRICS scalability on enterprise-scale knowledge graphs

Empirically validate the latency, memory usage, and practical retrieval and reasoning performance of the GRICS framework on substantially larger industrial or enterprise-scale knowledge graphs, including graphs containing millions of entities.

Background

The evaluation characterizes GRICS using the BRIDG-ICS knowledge graph, but the graph used in the study is not representative of substantially larger enterprise deployments. The paper notes that scaling may affect graph traversal, vector search, embedding maintenance, memory consumption, and operational overhead, particularly as graph density, update frequency, and reasoning depth increase.

Determining whether localized Cypher traversal and precomputed embedding retrieval remain efficient at enterprise scale is necessary for assessing the feasibility of deploying GRICS in large industrial environments and for identifying where approximate nearest-neighbour indexing, distributed graph processing, or alternative graph representations are required.

References

Although the current evaluation was conducted on the BRIDG-ICS knowledge graph, the architecture is designed to support larger industrial knowledge graphs, although its performance at substantially greater scale remains to be empirically validated.

— NeuroGraph: An AI Graph-Driven Neuro-Symbolic Framework for Explainable Threat Reasoning in Advanced Manufacturing  (2609.00604 - Nandiya et al., 1 Sep 2026) in Discussion, subsection “Scalability Considerations”; also Limitations and Open Research Questions, subsection “Limitations”

H3 and H4 remain untested under their falsification conditions.

— NS-ST-GraphRAG: Neuro-Symbolic Spatio-Temporal GraphRAG for Literary Knowledge Processing  (2609.05139 - Lin, 4 Sep 2026) in Section 5, subsection “Hypothesis-Wise Expectations”