metadata
license: mit
base_model:
- google-bert/bert-base-uncased
pipeline_tag: text-classification
tags:
- bbc-news-classification
- bert
BBC News Classification with BERT
Overview
This model classifies BBC news articles into five categories: Business, Entertainment, Politics, Sport, and Tech. It is fine-tuned using a pre-trained BERT model to achieve high accuracy in text classification.
Dataset
The model is trained on the BBC News dataset, which consists of categorized news articles. The dataset contains training and test splits.
- Train Dataset: Contains labeled news articles.
- Test Dataset: Contains unlabeled news articles for evaluation.
Model Details
- Base Model: bert-base-uncased
- Training Framework: PyTorch with Hugging Face Transformers
- Batch Size: 16
- Optimizer: Adam with Learning Rate Scheduling
How to Use
You can use this model to classify BBC news articles into one of the five categories. Below is an example using the Hugging Face Transformers library:
from transformers import pipeline
classifier = pipeline("text-classification", model="NotThareesh/BBC-News-Classifier-BERT")
labels = {"LABEL_0": "Business", "LABEL_1": "Entertainment", "LABEL_2": "Politics", "LABEL_3": "Sport", "LABEL_4": "Tech"}
text = "The stock market saw a significant rise today after tech companies reported high earnings."
result = classifier(text)
print(f"Predicted Label: {labels[result[0]['label']]}, Accuracy Score: {result[0]['score']}") # Predicted Label: Business, Accuracy: 0.996
Model Performance
- Accuracy: Achieved over 99% accuracy on the validation dataset.
- Evaluation: The model performs consistently on Precision, Recall, and F1-score.
License
This model is available under the MIT License.