---
configs:
- config_name: default
  data_files:
    - split: train
      path: "train/*.arrow"
    - split: test  
      path: "test/*.arrow"
---

[![CC BY-NC-SA 4.0][cc-by-nc-sa-shield]][cc-by-nc-sa]   [![CC BY-NC-SA 4.0][cc-by-nc-sa-image]][cc-by-nc-sa]

The **EduRABSA dataset** is licensed under a [Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License][cc-by-nc-sa].

[cc-by-nc-sa]: http://creativecommons.org/licenses/by-nc-sa/4.0/
[cc-by-nc-sa-image]: https://licensebuttons.net/l/by-nc-sa/4.0/88x31.png
[cc-by-nc-sa-shield]: https://img.shields.io/badge/License-CC%20BY--NC--SA%204.0-lightgrey.svg


# EduRABSA_ACD

This dataset is a version of the EduRABSA dataset formatted for the **ACD / AOC (Aspect Category Detection / Aspect-opinion Categorisation)** subtask.

Full details about the dataset are available in our paper [**EduRABSA: An Education Review Dataset for Aspect-based Sentiment Analysis Tasks**](
https://arxiv.org/abs/2508.17008
)

To cite this dataset:
```
@misc{hua2025edurabsaeducationreviewdataset,
      title={EduRABSA: An Education Review Dataset for Aspect-based Sentiment Analysis Tasks}, 
      author={Yan Cathy Hua and Paul Denny and Jörg Wicker and Katerina Taskova},
      year={2025},
      eprint={2508.17008},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2508.17008}, 
}
``` 


## Dataset Overview

**Education Review ABSA (EduRABSA)**  is a manually annotated student review text dataset for multiple Aspect-based Sentiment Analysis tasks, including:
* Aspect-(opinion-category)-Sentiment Quadruplet Extraction (ASQE)
* Aspect-(opinion)-Sentiment Triplet Extraction (ASTE)
* Aspect Sentiment Classification (ASC; a.k.a. Aspect Polarity Classification; APC)
* Aspect Category Detection (ACD) / Aspect-opinion Categorisation (AOC)
* Aspect-Opinion Pair-Extraction (AOPE)
* Aspect Extraction (AE)
* Opinion Extraction (OE)

The dataset consists of 6,500 pieces of stratified samples of public tertiary student review text in the English language released in 2020-2023 on courses ("course review", N=3,000), teaching staff ("teacher review", N=3,000), and university ("university review", N=500). 

## Details & Instructions


## Dataset Source

| **Review Type**   | **Dataset Name**                                                                                                 | **Publish Year** | **Licence**             | **Total Entries** | **Sampled (N=6,500)** |
|-------------------|------------------------------------------------------------------------------------------------------------------|------------------|--------------------------|-------------------|------------------------|
| Course review     | Course Reviews University of Waterloo [1]                                                                        | October 2022     | CC0: Public Domain       | 14,810            | 3,000                  |
| Teacher review    | Big Data Set from RateMyProfessor.com for Professors' Teaching Evaluation [2]                                    | March 2020       | CC BY 4.0                | 19,145            | 3,000                  |
| University review | University of Exeter Reviews [3]                                                                                 | June 2023        | CC0: Public Domain       | 557               | 500                    |

[1]: [Waterloo Course Reviews. *Course Reviews University of Waterloo*. October 2022](https://www.kaggle.com/datasets/anthonysusevski/course-reviews-university-of-waterloo).  
[2]: [RateMyProfessor Dataset. *Big Data Set from RateMyProfessor.com for Professors' Teaching Evaluation*. March 2020](https://data.mendeley.com/datasets/fvtfjyvw7d/2).  
[3]: [Exeter Reviews. *University of Exeter Reviews*. June 2023](https://www.kaggle.com/datasets/rohitpawar1/university-of-exeter-reviews).


## Sampling & Annotation Info
For sampling and annotation details, please visit the full repository: https://github.com/yhua219/edurabsa_dataset_and_annotation_tool