metadata
license: mit
language:
- en
tags:
- education
- movie
size_categories:
- 1K<n<10K
📚 Dataset Card for "My Sentiment Dataset"
Dataset Summary
This dataset contains English-language movie reviews labeled with sentiment (positive or negative). It can be used for binary sentiment classification tasks. The dataset was collected from publicly available online sources and manually labeled for accuracy.
- 10,000 reviews
- 5,000 positive / 5,000 negative
- Text only (no metadata)
Supported Tasks and Leaderboards
Task: Sentiment Classification
Input: Text (movie review)
Output: Binary label (positive or negative)
This dataset is commonly used for:
- Benchmarking binary text classification models
- Fine-tuning pretrained language models for sentiment analysis
Languages
- English (en)
Dataset Structure
Data Fields
text(string): The movie review text.label(int): Sentiment label — 0 for negative, 1 for positive.
Example
{
"text": "I really enjoyed this movie. The acting was great!",
"label": 1
}