Explore the dataset that served as the backbone for groundbreaking research in the automatic classification of Arabic poetry emotions using deep learning techniques. This dataset is a valuable resource for anyone delving into natural language processing, sentiment analysis, or Arabic literature studies.
- Comprehensive Dataset: The dataset includes 9,452 Arabic poems, meticulously labeled with emotions—Joy, Sadness, and Love—providing a robust foundation for training deep learning models.
- Structured Files:
poems.csv: Contains the poems along with their corresponding emotional labels.README.md: This file, offering detailed information about the dataset's contents and usage.
This dataset was utilized in the research paper titled "Classification of Arabic Poetry Emotions Using Deep Learning," published in the Computers journal in 2023. The study presents an in-depth analysis of various deep learning models applied to classify Arabic poetry into emotional categories.
- Shahriar, Sakib, Noora Al Roken, and Imran Zualkernan. "Classification of Arabic Poetry Emotions Using Deep Learning." Computers 12, no. 5 (2023): 89.
DOI: 10.3390/computers12050089
The automatic classification of poetry into categories such as author, era, or emotion is a challenging yet fascinating problem. Traditional approaches often rely on feature engineering and machine learning. However, this paper breaks new ground by applying deep learning techniques to classify Arabic poetry into emotional categories.
A new dataset of 9,452 Arabic poems labeled with emotions—Joy, Sadness, and Love—was developed for this study. Various deep learning models were trained on this dataset, including one-dimensional Convolutional Neural Networks (1DCNN), Gated Recurrent Unit (GRU), and Long Short-Term Memory (LSTM) networks, which achieved F1-scores of 0.62, 0.62, and 0.53, respectively.
Notably, the AraBERT model, an Arabic adaptation of the Bidirectional Encoder Representations from Transformers (BERT), outperformed other models with an accuracy of 76.5% and an F1-score of 0.77, setting a new state-of-the-art for Arabic poetry emotion classification.
If you use this dataset in your research or any other work, please cite the following paper:
Shahriar, Sakib, Noora Al Roken, and Imran Zualkernan. "Classification of Arabic Poetry Emotions Using Deep Learning." Computers 12, no. 5 (2023): 89.
DOI: 10.3390/computers12050089
Please acknowledge the authors by citing the reference above when using this dataset.
The dataset is made available under the terms specified in the license file.
For any further inquiries or questions about the dataset, please reach out to the authors of the paper.