Logo Lanfrica

Rif-Organization/Resources-Roadmap

Domaine:

natural language processing

Type de record:

project
Créateur:
Rif
Hôte:
All the Tarifiyt related resources that will be aggregated on short term. # Resources-Roadmap All the Tarifit related resources that will be aggregated for AI and software projects. ## Target Projects - **Tarifit Learning App** - Language learning application for Tarifit - **Translation Model** - Machine translation model for Tarifit ↔ other languages --- ## Resource Collection Checklist ### Text & Parallel Corpus Data - [ ] Curso de Lengua Tamazight Jahfar Hassan Yahia - [ ] Manual de Conjugación Rifeña v2.2 páginas con cubiertas ### Dataset resources for NLLB model - [x] Tarifit Bible with translations. - [x] Random resources (will elobare on this in future readme update) ### Audio & Speech Data ### Grammar & Linguistic Resources ### Learning App Specific Resources ### Translation Model Specific Resources --- ## Data Formats Resources should be compiled into the following formats for AI and software usability: | Resource Type | Recommended Format | Description | |---------------|-------------------|-------------| | Parallel corpus | `.tsv`, `.csv`, `.jsonl` | Tab/comma separated or JSON lines with source-target pairs | | Dictionary | `.json`, `.sqlite` | Structured key-value pairs with metadata | | Audio | `.wav`, `.mp3` | With corresponding transcription files | | Grammar rules | `.md`, `.yaml` | Human-readable with structured data | | Vocabulary lists | `.json`, `.csv` | Categorized word lists with translations | | Flashcards | `.json`, `.apkg` | Custom JSON or `.apkg` (Anki deck export format) | --- ## Resource Status Legend - [ ] Not started / Not collected - [ ] Collected and formatted - 🔄 In progress - ⚠️ Needs review/verification --- ## Contributing To contribute resources: 1. Ensure content is properly licensed or in the public domain 2. Format data according to the specifications above 3. Include metadata about source and quality 4. Submit via pull request with description of the resource --- ## Notes - Priority focus: Building parallel corpus for translation model - Secondary focus: Vocabulary …