Published Paper Details:

Modern criminal intelligence systems are heavily constrained by data silos and the structural fragmentation of heterogeneous records. Law enforcement operations generate vast streams of unstructured text narratives (e.g., field interrogations, witness accounts, surveillance summaries) alongside complex, highly connected relational data (e.g., asset ownership trails, deep familial lineages, judicial histories, vital status records, and hidden organizational backing matrices). Relational Database Management Systems (RDBMS) suffer from exponential performance degradation when calculating deep multi-hop network joins (>3 hops), while standalone Dense Vector Retrieval or Approximate Nearest Neighbor (ANN) vector spaces fail to capture explainable topological context, treating linked entities as isolated points. This paper introduces a highly scalable, single-system hybrid framework deployed entirely within an enterprise distributed property graph architecture (Nebula Graph). By embedding high-dimensional semantic vector properties directly into graph vertices and utilizing an inline, hardware-accelerated Cosine Similarity index, our framework eliminates the data synchronization latency, consistency overhead, and operational complexity inherent in decoupled multi-database pipelines. A Native Unified Vector-Relational Graph Intelligence Framework for Criminal Link Analysis and Explainable Evidence Board Generation. Furthermore, formalize the exact mathematical derivations for a unified multi-factor ranking equation balancing text semantic vector similarity, graph connectivity metrics, geographic alignment constants, and historical crime sentencing density. Finally, provide production-grade, executable nGQL (Nebula Graph Query Language) expressions and Python execution modules that deliver query-focused, explainable visual evidence boards. The empirical design framework demonstrates substantial capability improvements over standalone architectures, satisfying the strict requirements for modern digital forensics and court-admissible chain-of-custody tracking.