metadata
library_name: transformers
license: mit
datasets:
- Chidvi201/Twitter_Data.csv
language:
- en
metrics:
- accuracy
base_model:
- distilbert-base-uncased
pipeline_tag: text-classification
tags:
- sentiment-analysis
- transformers
- distilbert
- huggingface
- twitter
- airline
- gradio
- text-classification
✈️ BERT Airline Sentiment Classifier
A DistilBERT-based model fine-tuned for sentiment analysis on airline-related tweets. It classifies input text into positive, neutral, or negative sentiment categories.
🧠 Model Details
Model Description
This model uses distilbert-base-uncased as a base and is fine-tuned on a cleaned dataset of airline tweets. It performs multi-class classification with 3 sentiment labels.
- Developed by: Nicolettem
- Model type: Transformer-based (DistilBERT)
- Language(s): English (
en) - License: MIT
- Fine-tuned from:
distilbert-base-uncased
📦 Model Sources
- Repository: https://huggingface.co/Nicolettem/bert-sentiment-nic
- Demo Space: Gradio Demo
- Dataset: Chidvi201/Twitter_Data.csv
💡 Uses
Direct Use
You can use this model to classify sentiment of customer reviews or tweets — especially in the airline or travel domain.
Downstream Use
It can serve as a base for training more domain-specific sentiment models, or be integrated into social media monitoring tools.
Out-of-Scope Use
- The model was not trained on non-English tweets or formal customer service chat.
- It may reflect dataset biases and perform poorly on sarcastic or ambiguous text.
⚠️ Bias, Risks, and Limitations
- May reflect biases in public Tweets