The AnubisX Framework is a comprehensive scientific methodology for determining the identity of a human operator from their digital behavioral patterns. It defines a complete theoretical specification including 31 formal axioms, 37+ algorithm specifications, 292 mathematical definitions, 50 equations, 38 experiment designs, 24 benchmarks with 30 baselines, 20 case studies, and a four-tier validation framework with 33 quantified acceptance criteria. The framework includes a validated prototype implementation (Anubis Twitter v2.5) demonstrating stylometric fingerprinting with 372-dim feature vectors and FAISS-based similarity search at 16μs, supported by 15 executed experiments on 31 Egyptian Twitter accounts. The framework shifts digital attribution from transient technical artifacts (IP addresses, device fingerprints) to persistent human cognitive signatures rooted in stable cognitive processing habits.

If you use the AnubisX Framework in your research, please cite this repository.