roudik/Brand_Classification_Armenian
A BiLSTM-based text classification system for mapping Armenian/Latin retail product names to ADG codes (brand categories), with evaluation, inference, and a lookup tool. The project includes data preprocessing, model training, and a full pipeline for a specific multi-class NLP task.
AzgIntel is an AI-powered NLP pipeline for Armenian text classification using a BERT-based multilingual model. It pro...
A demonstration repository for a pre-trained Armenian text embedding model, showcasing applications in text classific...
A machine learning project for classifying handwritten Armenian letters using the Mashtots Dataset v2 from Kaggle. It...
A machine learning model for classifying Armenian letters (both uppercase and lowercase) using a Convolutional Neural...
A small NLP project that trains classifiers to distinguish between Eastern and Western Armenian dialects using Wikipe...
A curated multilingual dataset of Armenian and Armenia-related keywords, names, and geographic terms designed for fil...