Pericles001/augmentation_tasks_fon_ha
Domaine:
natural language processingType de record:
project
Créateur:
Per
Hôte:
Project to run augmentation and NER/POS tasks on hausa and fongbe languages
# Data Augmentation for Low-Resource African NLP
**Comparing LLM-based Augmentation vs Back-Translation for Hausa and Fongbe**
This project investigates whether synthetic data augmentation improves NLP model performance for low-resource African languages. We compare two augmentation strategies across three tasks: Named Entity Recognition (NER), Part-of-Speech (POS) tagging, and Sentiment Analysis.
---
## Table of Contents
- Research Questions
- Languages & Hypothesis
- Tasks & Datasets
- Project Structure
- Installation
- Quick Start Guide
- Running Experiments
- Expected Results
- Troubleshooting
- Citation
---
## Research Questions
1. **RQ1**: Does data augmentation improve NER/POS/Sentiment performance for Hausa and Fongbe?
2. **RQ2**: Which method works better: LLM-generated synthetic data or back-translation?
3. **RQ3**: Does the effectiveness vary by language and task?
---
## Languages & Hypothesis
| Language | ISO Code | Translation Quality* | Expected Outcome |
|----------|----------|---------------------|------------------|
| **Hausa** | `hau` | High (4.46/5) | Augmentation should **help** |
| **Fongbe** | `fon` | Low (2.20/5) | Augmentation may **hurt or be neutral** |
*Based on prior SIGIR paper findings on LLM translation quality for African languages.
**Key Hypothesis**: High translation quality → augmentation helps; Low quality → may introduce noise.
---
## Tasks & Datasets
| Task | Dataset | Metric | Train Size (Hausa/Fongbe) |
|------|---------|--------|---------------------------|
| **NER** | MasakhaNER 2.0 | F1 Score | 5,716 / 4,343 |
| **POS** | MasakhaPOS | Accuracy | 753 / 810 |
| **Sentiment** | AfriSenti | F1 Weighted | 14,172 (Hausa only) |
### NER Labels
`O, B-PER, I-PER, B-ORG, I-ORG, B-LOC, I-LOC, B-DATE, I-DATE`
### POS Labels (Universal Dependencies)
`NOUN, VERB, ADJ, ADV, PROPN, PRON, DET, ADP, NUM, CONJ, PUNCT, ...`
### Sentiment Labels
`positive, negative, neutral`
---
## Experimental Conditions
| Condition | Des …