File size: 2,928 Bytes
0e1fab3 | 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 | ---
language:
- ha
license: apache-2.0
task_categories:
- summarization
---
# Dataset Card for Dataset Name
<!-- Provide a quick summary of the dataset. -->
Hausa text-extractive ATS evaluation dataset. The dataset comprises 113 Hausa news articles from different genres, including sports, religion, politics, and culture. For each news article, there are two corresponding, manually generated gold standard summaries, whose lengths are 20% of the original article.
- **Curated by:** Abdulkadir Abubakar Bichi
- **Language(s) (NLP):** Hausa
- **Repository:** (https://doi.org/10.1371/journal.pone.0285376.s001)
- **Paper:** [[More Information Needed]](https://doi.org/10.1371/journal.pone.0285376)
### Source Data
<!-- This section describes the source data (e.g. news text and headlines, social media posts, translated sentences, ...). -->
The dataset comprises 113 Hausa news articles from different genres, including sports, religion, politics, and culture.
#### Data Architecture
<!-- This section describes the data collection and processing process such as data selection criteria, filtering and normalization methods, tools and libraries used, etc. -->
Each entry in the dataset contains the following fields:
id: a unique string identifier for each example.
article: a list[string] field representing the original news article.
refrence1: a list[string] field representing the professionally gold summary of the article.
#### Usage
<!-- This section describes the people or systems who originally created the data. It should also include self-reported demographic or identity information for the source data creators if this information is available. -->
The extractive dataset can be used to tevaluate models for extractive text summarization tasks on Hausa Language single documents.
It allows models to learn to predict which sentences from an original text contribute to a summary.
The 'reference1 and reference2' field can serve as a basis for comparison, helping to assess how well the selected sentences cover the key
points in the article.
#### Who are the annotators?
<!-- This section describes the people or systems who created the annotations. -->
Abdulqahar M. Abubakar and Abdulaziz Aminu
## Citation [optional]
<!-- If there is a paper or blog post introducing the dataset, the APA and Bibtex information for that should go in this section. -->
**BibTeX:**
@article{Bichi2023GraphbasedET,
title={Graph-based extractive text summarization method for Hausa text},
author={Abdulkadir Abubakar Bichi and Ruhaidah Samsudin and Rohayanti Hassan and Layla Rasheed Abdallah Hasan and Abubakar Ado Rogo},
journal={PLOS ONE},
year={2023},
volume={18},
url={https://api.semanticscholar.org/CorpusID:258587667}
}
**APA:**
[Bichi, A.A., Samsudin, R., Hassan, R., Hasan, L.R., & Ado Rogo, A. (2023). Graph-based extractive text summarization method for Hausa text. PLOS ONE, 18.]
|