# Short Text Analysis

BERT-Emotion is a lightweight natural language processing model based on bert-lite and NeuroBERT-Mini, optimized for short-text emotion detection on edge and IoT devices.

A named entity recognition model fine-tuned based on electra-small, capable of identifying three types of entities: locations, person names, and organizations

This model is fine-tuned based on the BERT architecture for detecting gibberish text (e.g., random character combinations) in Turkish.

A fine-tuned DistilBERT-based model for classifying COVID-related tweets

A text classification model based on BERT-mini architecture for identifying purchase or sales intent in text

A Chinese text classification model based on the DistilBERT architecture, trained on 4 categories: ['Women', 'Sports', 'Literature', 'Campus'].