File size: 3,925 Bytes
7b50f29 da422cf 7b50f29 9e5b5ca 7b50f29 0344de0 7b50f29 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
---
language:
- ar
tags:
- multi label text-classification
- arabic
- app-reviews
- nlp
dataset_info:
features:
- name: review
dtype: string
- name: appname
dtype: string
- name: platform
dtype: string
- name: judg_one
dtype: string
- name: judg_two
dtype: string
- name: judg_three
dtype: string
- name: judg_four
dtype: string
- name: judg_five
dtype: string
splits:
- name: train
num_examples: 2900
license: mit
---
# AURA-Classification (Multi-Label Version)
## Dataset Description
The AURA (App User Review in Arabic) Classification dataset is a collection of 2,900 Arabic-language app reviews collected from various mobile applications. This dataset is designed for multi-label text classification, where each review can belong to multiple classes simultaneously.
Each review in the dataset was independently annotated by five different annotators. To construct the multi-label version of the dataset, a review is assigned to a given class if at least one annotator labeled it with that class. As a result, a single review may be associated with up to four labels.
### Features
The dataset includes the following fields:
- **review**: The text of the review in Arabic.
- **appName**: The name of the application being reviewed.
- **platform**: The platform (iOS or Android) where the review was posted.
- **judg_one, judg_two, judg_three, judg_four, judg_five**: The labels assigned by each of the five annotators.
The possible classification labels are:
- `bug_report`: The review highlights a bug or issue in the app.
- `improvement_request`: The review suggests improvements or features.
- `rating`: The review expresses a general rating or opinion.
- `others`: Miscellaneous or uncategorized reviews.
### Dataset Statistics
- **Total Reviews**: 2,900
- **Platforms**: iOS, Android
- **Applications**: Multiple apps from diverse categories.
- **Labels Distribution**: Four possible classes in a multi-label setting (each review may have multiple labels).
### Example Entry
```json
{
"review": "الرجاء تحديث التطبيق لانه اذا سويت البلاغ في النهاية يخرج من الطبيق ولا يتم
إرسال البلاغ",
"appname": "تقديم بلاغ مخالفة تجارية",
"platform": "android",
"judg_one": "bug_report",
"judg_two": "improvement_request",
"judg_three": "bug_report",
"judg_four": "improvement_request",
"judg_five": "bug_report"
}
```
## Use Cases
This dataset is suitable for:
- Multi-label text classification of app reviews.
- Issue identification and requirements elicitation.
- Multilingual NLP research focused on Arabic.
- Fine-tuning models for app review classification.
## Citation
If you use this dataset, please cite it as:
```
@article{Aljeezani2025arabic,
title={Arabic App Reviews: Analysis and Classification},
author={Aljeezani, Othman and Alomari, Dorieh and Ahmad, Irfan},
journal={ACM Transactions on Asian and Low-Resource Language Information Processing (TALLIP)},
volume={24},
number={2},
pages={1--28},
year={2025},
publisher={ACM New York, NY, USA},
}
@article{alansari2025multilabel,
title = {Multi-Label Classification of Arabic App Reviews with Data Augmentation and Explainable AI},
author = {Alansari, Aisha and Alomari, Dorieh and Mahmood, Sajjad and Ahmad, Irfan},
journal = {Arabian Journal for Science and Engineering},
year = {to-appear},
publisher = {Springer}
}
```
## License
This dataset is shared under the [MIT License](https://opensource.org/licenses/MIT). Please ensure appropriate attribution when using this dataset.
## Acknowledgments
Special thanks to the contributors and reviewers who made this dataset possible.
## Contact
For questions or feedback, please reach out to the corresponding author (Irfan Ahmad).
|