Overview#
Argus Entity Resolution and Relationship Graph Intelligence delivers knowledge graph capabilities combining advanced entity resolution, relationship mapping, and network analysis to expose hidden connections across millions of entities and relationships at unlimited degrees of separation. The platform enables intelligence agencies, law enforcement, financial crimes units, and corporate investigators to discover complex criminal networks, expose hidden ownership structures, map organisational hierarchies, and uncover sophisticated fraud schemes through graph-native analytics.
AI-powered entity resolution, real-time relationship inference, and advanced graph algorithms including PageRank, community detection, and betweenness centrality transform disconnected data points into actionable intelligence. The system automatically identifies duplicate entities across sources, infers hidden relationships through pattern analysis, and visualises networks with interactive rendering of large-scale graphs.
The platform scales from small investigative queries to enterprise-wide intelligence analysis, handling millions of entities and relationships while maintaining sub-second query performance for even complex multi-hop traversals.
Key Features#
Entity Management#
- Large-scale entity management across person, organisation, location, asset, vehicle, and account profiles
- Comprehensive entity profiles with attributes, aliases, identifiers, and source tracking
- Entity lifecycle management including creation, merging, splitting, and archival
- Multi-source data fusion combining attributes from disparate systems into unified entity records
Relationship Mapping#
- Multi-type relationship connections with temporal tracking and weighted edges
- Bidirectional and directional relationship support for accurate network modelling
- Temporal relationship tracking showing how networks evolve over time
- Relationship strength scoring based on frequency, recency, and corroborating evidence
- Automated relationship inference discovering hidden connections through pattern analysis
- Centrality scoring identifying the most influential or connected entities within criminal networks
- Shortest path analysis revealing the most efficient connections between entities of interest
Entity Resolution#
- AI-powered entity disambiguation with explainable merge decisions and high resolution accuracy
- Cross-source entity resolution with configurable matching rules and confidence scoring
- Phonetic matching, fuzzy string comparison, and attribute-based similarity analysis
- Human-in-the-loop review workflows for ambiguous matches requiring analyst judgment
- Continuous background resolution processing new records against the existing entity corpus
Graph Analytics#
- Advanced graph algorithms including PageRank, Louvain community detection, and betweenness centrality
- Shortest path analysis for determining connection routes between entities of interest
- Community detection identifying clusters of closely related entities within larger networks
- Influence analysis ranking entities by network centrality and connectivity measures
- Anomaly detection identifying unusual patterns in entity relationships and behaviours
Visualisation#
- Interactive network visualisation with force-directed layouts, temporal animations, and filtering
- Configurable graph rendering with entity type icons, relationship labels, and attribute display
- Timeline view showing network evolution and relationship formation over specified periods
- Geographic overlay mapping entity relationships onto spatial coordinates
- Subgraph extraction isolating relevant portions of large networks for focused analysis
- Print-ready visualisation export for inclusion in briefings and court documentation
- Export of visualisations for inclusion in reports and prosecution packages
- Graph comparison tools enabling analysts to examine structural changes in networks over time
- Cluster detection algorithms automatically identifying tightly connected subgroups within larger networks
Use Cases#
Criminal Network Analysis. Map the full organisational structure of criminal enterprises, identify key players through centrality analysis, discover communication patterns that reveal leadership hierarchies and operational cells, and track network evolution over time to anticipate organisational changes.
Financial Crime Investigation. Trace complex money laundering networks, expose shell company structures, identify beneficial ownership chains, and map relationships between accounts, entities, and transactions across multiple financial institutions and jurisdictions.
Intelligence Fusion. Resolve and connect entities across disparate intelligence sources, automatically discovering relationships and patterns that would be invisible through manual analysis. Reduce analyst workload by surfacing the most significant connections from massive datasets.
Fraud Ring Detection. Apply community detection algorithms to identify clusters of related entities involved in coordinated fraud, insurance schemes, or identity theft operations. Visualise the full scope of fraud networks to support comprehensive investigation and prosecution.
Integration#
- Connects with investigation and case management systems for seamless workflow integration
- Integrates with financial transaction monitoring platforms for real-time relationship enrichment
- Links to external intelligence feeds and watchlist databases for entity enrichment
- Supports export of network analysis results for reporting and prosecution packages
- Works with alert and notification systems for real-time threat detection and network change monitoring
- Compatible with geospatial analysis platforms for location-aware network visualisation
- Feeds into analytical dashboards for organisational intelligence oversight
- Pattern template library with pre-built network patterns for common criminal organisations
- Multi-hop query optimisation for rapid traversal across deep relationship chains
Open Standards#
- STIX 2.1 / TAXII 2.1 (OASIS): Entities and relationships are bidirectionally converted between the internal model and STIX 2.1 Structured Threat Information eXpression SDOs and SROs; an async TAXII 2.1 polling client enables automated ingestion from external threat intelligence feeds.
- GraphQL (June 2018 Specification): All entity queries, relationship traversals, graph statistics, and real-time subscription updates are exposed through a typed GraphQL API, enabling flexible client-driven data fetching.
- openCypher (Neo4j Cypher Query Language): Entity and relationship graph traversals against the Neo4j-backed property graph use the openCypher MATCH/RETURN pattern, with path-finding and multi-hop neighbourhood queries expressed in Cypher.
- W3C PROV-DM (Provenance Data Model): Entity merge decisions are recorded as provenance events conforming to the W3C PROV-DM specification, providing an auditable lineage trail for all entity resolution operations.
- Privacy-Preserving Record Linkage (PPRL), Bloom-filter / Dice coefficient (Schnell et al. 2009): Entity resolution encodes personal attributes as Bloom filters and computes Dice-coefficient similarity scores to support privacy-safe cross-source deduplication without exposing raw PII.
- ISO 8601: All temporal attributes on entities, relationships, and merge operations, including edge start/end dates and operation timestamps, are serialised as ISO 8601 datetime strings throughout the graph model.
- RFC 4122 (UUID): Every entity node and graph relationship is assigned a RFC 4122 Version 4 universally unique identifier, ensuring globally collision-resistant identity across multi-source data fusion.
Last Reviewed: 2026-02-23 Last Updated: 2026-04-14