Addisu22/Amharic_Ecommerce_Extractor

Building an Amharic E-commerce Data Extractor Build a Complete Amharic E-commerce Data Extractor ================================================== The project encompasses several key objectives, starting with the collection and preprocessing of Amharic data. Following this, the focus will shift to labeling the Amharic data specifically for Named Entity Recognition (NER). Subsequently, existing models will be fine-tuned to enhance their performance in NER tasks. Finally, a comparative analysis will be conducted to evaluate the effectiveness of various NER models. Task 1: Data Ingestion and Data Preprocessing ============================================== amharic_ecommerce_extractor │ ├── main.py               # Main script to run extractor ├── config.py               # API credentials & settings ├── preprocess.py       # Amharic text cleaning, tokenizing ├── extractor.py            # Telegram ingestion logic ├── utils.py                   # Helpers: saving, logging, formatting ├── data/ │ └── messages.csv       # Output structured messages ├── media/ │ ├── photos/               # Downloaded images │ └── docs/               # Downloaded documents ├── requirements.txt Task 2 : Label a Subset of Dataset in CoNLL Format ================================================== Task 3: Fine Tune NER Model =========================== Task 4: Model Comparison & Selection ==================================== Task 5: Model Inte …