---
library_name: transformers
tags:
- bist30
- turkish
- bert
- twitter
- ner
language:
- tr
base_model:
- dbmdz/bert-base-turkish-128k-uncased
---

# 🏷️ Financial NER Model for Entity Extraction

This is a fine-tuned BERT-based token classification model that extracts **structured financial entities** from Turkish tweets related to the stock market.

Entities include:
- 🏢 Stock Tickers (e.g., `AKBNK`, `EREGL`)
- 🎯 Target Prices (`Hedef`)
- 📉 Support Prices (`Destek`)
- 📈 Resistance Prices (`Direnç`)
- ⏱ Time spans (`Vade`)
- 📊 Predicted values (`Tahmin`)

It is a core part of a larger NLP pipeline for parsing and analyzing finance-related predictions on social media.

---

## 🧠 Model Details

- **Developed by:** damlakonur
- **Model type:** `BERT` fine-tuned for `token-classification`
- **Language(s):** Turkish
- **Finetuned from:** `bert-base-cased`
- **Trained using:** Hugging Face `Trainer` API
- **License:** MIT

---

## 🚀 How to Use

```python
from transformers import pipeline

model = pipeline(
    "token-classification",
    model="your-username/financial-ner-entities-bist30",
    aggregation_strategy="simple"
)

text = "#AKBNK hedef 60 TL, destek 52 TL."
output = model(text)
print(output)