DamiAIExpert/A-Standard-Hausa-Syllabifier

# A Standard Hausa Syllabifier Rule-based Hausa syllabification system with a pilot gold-standard dataset, evaluation script, manuscript figures, and supplementary materials for the paper: **Hausa Syllable Structure and Rule-Based Syllabification: A Linguistic and Computational Analysis** ## What Is Included - `syllabifier.py`: transparent rule-based Hausa syllabification engine. - `app.py` and `templates/index.html`: Flask web interface. - `data/hausa_syllabification_gold_pilot.csv`: 91-item pilot gold-standard set. - `evaluate_syllabifier.py`: pilot conformance evaluation against the gold set. - `ANNOTATION_GUIDELINES.md`: protocol for expanding the dataset. - `tools/build_figures.py`: regenerates manuscript figures. - `tools/build_revised_manuscript.py`: regenerates the revised manuscript files. - `figures/`: generated manuscript figures. - `A_Rule-Based_Hausa_Syllabifier_JWAL_REVISED.docx`: revised manuscript draft. - `A_Rule-Based_Hausa_Syllabifier_JWAL_REVISED.md`: Markdown manuscript draft. ## Install ```bash python -m venv .venv .venv\Scripts\activate pip install -r requirements.txt ``` ## Run the Web App ```bash python app.py ``` Then open: ```text 127.0.0.1 ``` ## Evaluate ```bash python evaluate_syllabifier.py ``` The current 91-item dataset is a curated pilot set. Its scores should be read as a pilot conformance check, not as broad Hausa lexical accuracy. ## Regenerate Figures and Manuscript ```bash python tools/build_revised_manuscript.py ``` This regenerates the manuscript, Markdown draft, cover letter, and figures. ## Data Note Raw Hausa text files used for candidate extraction are not included in this repository because they may contain third-party source material. Public releases should contain derived word forms, syllabification labels, metadata, and annotation decisions unless raw-text redistribution rights are clear.