README.md

metadata

license: apache-2.0
task_categories:
  - sentence-similarity
language:
  - ar
tags:
  - sentence-transformers
size_categories:
  - 1K<n<10K

Arabic STSB Structure

• The Arabic Version of the the Semantic Textual Similarity Benchmark (Cer et al., 2017)
• it is a collection of sentence pairs drawn from news headlines, video and image captions, and natural language inference data.
• Each pair is human-annotated with a similarity score from 1 to 5. However, for this variant, the similarity scores are normalized to between 0 and 1.

Examples:

{
  "sentence1": "طائرة ستقلع",
  "sentence2": "طائرة جوية ستقلع",
  "score": 1.0
}

{
  "sentence1": "رجل يعزف على ناي كبير",
  "sentence2": "رجل يعزف على الناي.",
  "score": 0.76
}

Collection strategy:

• Reading the sentences and score from the STSB dataset and dividing the score by 5.
• Deduplified: No

Disclaimer

Please note that:

• the translated sentences are generated using neural machine translation and may not always convey the intended meaning accurately.
• the similarity scores are normalized, and the original scores were between 1 and 5.

Contact

Contact Me if you have any questions or you want to use thid dataset

Note

Original work done by SentenceTransformers

Citation

If you use the Arabic Matryoshka Embeddings Dataset, please cite it as follows:

@misc{nacar2024enhancingsemanticsimilarityunderstanding,
      title={Enhancing Semantic Similarity Understanding in Arabic NLP with Nested Embedding Learning}, 
      author={Omer Nacar and Anis Koubaa},
      year={2024},
      eprint={2407.21139},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2407.21139}, 
}