aelhaloui-del/Amazigh-Philologia

I decipher Old Amazigh inscriptions # Amazigh-Philologia Research tools and computational implementations for Amazigh philology and epigraphy. ## EMP (Exclusion Matrix Protocol) Computational implementation of the Exclusion Matrix Protocol for segmenting Libyco-Amazigh scriptura continua inscriptions. ### Overview EMP generates Complete Lexical Segmentation Sets (CLSS) through: - **Tier 1:** Bare root validation against cross-dialectal lexicon (705 Hdaddou roots) - **Tier 2:** Morphological pattern recognition (9 productive Amazigh patterns) - **MI Ranking:** Monoconsonantal Index for plausibility assessment ### Installation No external dependencies required. Uses Python 3.8+ standard library only. ```bash git clone github.com cd Amazigh-Philologia ``` ### Quick Start **Basic Usage:** ```python from emp_tier2 import run_emp # Analyze an inscription inscription = "WSKWSRZMNMRW" # IKKIS 1 Line 2 result = run_emp(inscription) print(result) ``` **Command Line:** ```bash python emp_tier2.py.py # Enter inscription when prompted ``` ### Repository Contents **EMP Implementation:** - `emp_tier1.py.py` - Root-only segmentation engine (Tier 1 validation) - `emp_tier2.py.py` - Two-tier segmentation with morphological patterns ### Algorithm EMP implements a three-stage protocol: 1. **Partition Generation** - Binary boundary encoding (2^(n-1) partitions) 2. **Two-Tier Validation** - Lexical validation via bare roots + morphological decomposition 3. **MI Classification** - Non-exclusionary ranking by monoconsonantal index **Performance:** - n=6 (CALE3_1): 32 partitions, <0.1 seconds - n=12 (IKKIS 1): 2,048 partitions, ~2 seconds - Practical limit: n ≈ 20 graphemes ### Citation If you use this code in your research, please cite: ```bibtex @article{elhaloui2026emp, author = {El Haloui, Abdellah}, title = {Exclusion Matrix Protocol for Libyco-Amazigh Inscription Segmentation}, journal = {[To be updated upon publication]}, year = {2026}, note = {Code: http …