Graph Traversal Algorithms

Overview#

A blockchain forensics team receives a tip: funds from a known ransomware wallet have been moved. The wallet's balance cleared three days ago. Where did it go? The answer requires traversing a transaction graph with millions of nodes, following every possible path the funds could have taken, eliminating dead ends, and surfacing paths that terminate at exchange deposit addresses. The traversal algorithms process more than a million nodes per second. What would take a human analyst days of manual tracing takes the system seconds.

The Graph Traversal Algorithms module delivers high-performance graph navigation capabilities through optimised implementations of breadth-first search, depth-first search, and advanced pathfinding algorithms. It supports investigations across financial forensics, criminal network analysis, and intelligence operations, enabling analysts to discover hidden connections, trace fund flows, and map complex relationships across massive graphs built on the Neo4j analysis layer.

Key Features#

• Eight traversal algorithm primitives: breadth-first search, depth-first search, Dijkstra shortest path, A* heuristic search, all paths enumeration, bidirectional search, depth-limited exploration, and custom predicate-based traversal
• High-throughput traversal processing exceeding one million nodes per second on standard infrastructure
• Breadth-first search with level-order exploration guaranteeing shortest path discovery in unweighted graphs
• Depth-first search excelling at cycle detection, topological sorting, path enumeration, and strongly connected component identification
• Shortest path algorithms with guaranteed optimality for non-negative weights and heuristic-guided acceleration
• K-shortest paths discovery providing ranked alternative routes with diversity scoring
• All paths enumeration revealing every possible connection route between entities for comprehensive analysis
• Bidirectional search simultaneously exploring from both endpoints for significant speedup on large graphs
• Depth-limited search preventing infinite recursion with controlled exploration depth and iterative deepening
• Custom predicate traversal enabling domain-specific path filtering through user-defined boolean conditions
• Parallel execution across independent subgraphs with optimised queue management
• Bloom filter optimisation reducing memory overhead for visited set tracking
• Early termination and incremental result streaming for responsive query processing
• Configurable timeouts preventing runaway queries in complex graph topologies

Use Cases#

• Blockchain Fund Flow Tracing: Forensics teams trace cryptocurrency movements from known illicit sources through multi-hop transaction networks with configurable value thresholds
• Criminal Network Proximity Analysis: Law enforcement identifies all entities within specified hops of a suspect for comprehensive investigation targeting
• Attack Path Discovery: Cybersecurity teams enumerate all potential lateral movement paths from compromised hosts to critical assets for vulnerability assessment
• Supply Chain Dependency Mapping: Organisations trace complete supply routes with required waypoint constraints and transit time optimisation

Integration#

• Connects with the Neo4j graph analysis layer through typed APIs with real-time query processing
• Compatible with investigation platforms for automated network exploration and connection discovery
• Supports WebSocket-based result streaming for progressive delivery of large traversal results
• Node-level security ensuring traversals respect row-level access policies
• Complete audit logging of all traversal queries with user context
• Multi-tenant graph isolation with automatic query scoping
• Compliance with GDPR, CCPA, and HIPAA through automatic PII filtering and encryption

Open Standards#

• GraphQL (June 2018 Specification): All traversal queries, mutations, and real-time subscriptions are exposed through a GraphQL API, enabling typed, self-documenting access to pathfinding and community-detection operations.
• graphql-transport-ws Protocol: Incremental traversal results and live graph update subscriptions are streamed over WebSocket using the graphql-transport-ws subprotocol, providing ordered, multiplexed delivery of large result sets.
• RFC 6455 (WebSocket): The underlying full-duplex WebSocket connection carries the graphql-transport-ws subprotocol, enabling low-latency progressive delivery of traversal output without polling.
• SQL:1999 Recursive CTEs (ISO/IEC 9075 WITH RECURSIVE): All BFS, DFS, shortest-path, and depth-limited traversals execute as ISO-standard recursive common table expressions over the PostgreSQL relationship store, with no proprietary graph-database extension required.
• GEXF 1.3 (Graph Exchange XML Format): Traversal result sets can be exported in GEXF 1.3, the open XML format maintained by the GEXF Working Group, for direct import into tools such as Gephi and Cytoscape.
• JSON (RFC 8259): Node and edge property payloads, traversal parameters, and streaming result objects are all serialised as RFC 8259 JSON throughout the API and persistence layer.
• OAuth 2.0 (RFC 6749) / OpenID Connect: Every traversal endpoint enforces bearer-token authentication issued via the platform's OAuth 2.0 / OIDC authorisation server, ensuring all graph queries carry a verified identity and tenant scope.

Last Reviewed: 2026-02-23 Last Updated: 2026-04-14