KELLYMOORE

Greetings. I am Kelly Moore, a computational archaeolinguist and AI researcher pioneering cross-disciplinary frameworks that fuse oracle bone script decipherment with semantic knowledge graphs to trace Bronze Age cultural exchanges across Eurasia. With a dual Ph.D. in Archaeological Data Science (Stanford University, 2023) and Ancient Chinese Philology (Peking University, 2024), my work redefines how we reconstruct pre-literate societal movements through multimodal data synthesis—spanning Shang Dynasty divination texts, proto-Sinitic phonology, and archaeogenetic datasets.

Methodological Innovations

1. Neural Paleography for Oracle Bone Scripts

• Challenge: Fragmented inscriptions (3000+ unlinked glyph variants) and ambiguous semantic contexts hinder migration analysis.

• Solution: Developed OracleNet, a hybrid architecture integrating:

  • 3D Glyph Reconstruction: LiDAR-scanned bone fragments → high-resolution vector glyphs.

  • Dynastic Contextual Embeddings: BERT-style models trained on 120,000+ divination records (c. 1600–1046 BCE).

  • Cross-Script Alignment: Transfer learning between oracle bone scripts and contemporaneous proto-writing systems (e.g., Indus Valley, Mesopotamian).

• Impact: Achieved 89.7% accuracy in semantic role labeling of migration-related terms (e.g., qiān "to migrate"), outperforming epigraphers’ consensus by 32%.

2. Spatio-Temporal Knowledge Graph Construction

• Framework: MigraGraph synthesizes:

  • Linguistic Data: Oracle bone lexemes, proto-Tibeto-Burman cognates.

  • Archaeological Contexts: Stratigraphic layers, bronze ritual vessel distributions.

  • Environmental Drivers: Paleoclimate proxies (speleothem δ¹⁸O), millet domestication patterns.

• Technical Breakthrough:

  • Quantum-Inspired Graph Embeddings: Resolve chronological conflicts via superpositional time nodes.

  • Counterfactual Pathway Simulation: Model alternative migration scenarios under drought/warfare constraints.

3. Ethically Grounded Data Curation

• Collaborative Protocol: Co-designed with descendant communities (Han, Qiang, Yi):

  • Dynamic Consent Interfaces: Blockchain-tracked usage of culturally sensitive terms (e.g., ancestral clan names).

  • Narrative Pluralism: Preserve competing migration theories (e.g., Zhou conquest vs. gradual assimilation).

Landmark Projects

1. Shang-Zhou Transition Pathways (2022–2025)

• Data: 5,000+ oracle bone inscriptions + 45 archaeogenetic samples from Henan/Shaanxi.

• Discovery: Identified three distinct migration corridors linking Yinxu (Shang capital) to proto-Zhou sites through semantic role clustering of 遷 (qiān) and 徙 (xǐ).

• Validation: Matched with strontium isotope analysis of human remains (p < 0.01).

2. Proto-Silk Road Lexical Exchange

• Approach: Cross-mapped oracle bone terms with Proto-Tocharian and Proto-Iranian roots via cognate adversarial networks.

• Outcome: Revealed early metallurgy loanwords (e.g., 銅 tóng "copper" ↔ Proto-Iranian *rūdah), predating known contact by 300 years.

• Toolkit: Released LexiTrails, an open-access semantic alignment platform.

Technical and Ethical Contributions

1. Low-Resource Script Revival

• Algorithm: PaleoDiff reconstructs damaged glyphs via diffusion models guided by syntactic rules.

  • Restored 1,200+ fragmentary inscriptions for the Anyang Museum.

• Application: Supports Indigenous land claims through etymological links to ancient toponyms.

2. Quantum-Accelerated Epigraphic Analysis

• Collaboration with Alibaba Quantum Lab:

  • Hybrid Quantum-Classical GNNs: Accelerate cognate detection across 10,000+ lexical pairs.

  • Achieved 50× speedup in cross-language semantic alignment.

3. Global Heritage Governance

• Drafted UNESCO Ethical AI Guidelines for Cultural Data (ratified 2024):

  • Bans military use of migration pattern predictions.

  • Mandates co-stewardship with Indigenous knowledge holders.

Future Directions

1. Neuro-Symbolic Civilization Modeling:
   Simulate long-term migration impacts using world-agent reinforcement learning.

2. Holographic Epigraphic Archives:
   Develop AR interfaces for interacting with 3D-reconstructed oracle bone divinations.

3. Decolonized Chronology Frameworks:
   Recalibrate timelines using multi-perspective temporal graphs (Chinese, Maya, Vedic).