TeSum: Telugu Abstractive Summarization

[anoopkunchukuttan]

TeSum: Human-Generated Abstractive Summarization Corpus for Telugu

Link to dataset broken - https://github.com/manshri/TeSum/

[ashokurlana]

Please find the TeSum dataset: https://ltrc.iiit.ac.in/showfile.php?filename=downloads/teSum/

[anoopkunchukuttan]

Thanks, I noticed that the download file is named as a zip, but is actually a tgz archive.

[anoopkunchukuttan]

• Telugu Abstractive Summarization dataset containing 20k+ article-summary pairs (with a 80,20,20% train/valid/test split). That's a fairly big, manually created dataset.
• The summaries were collected via a crowdsourced effort with detailed guidelines and quality control to ensure high quality.
• Unlike other automatically mined summarization datasets, which pair article with highlights as summary, the summaries are manually written.
• 90k summaries were thus collected, and finally 20k were accepted after quality control. The scale and co-ordinaltion of the effort is laudable.
• The paper documents the guidelines and process in detail and will be a useful reference for other interested in summarization dataset creation guidelines.

Paper Link: https://aclanthology.org/2022.lrec-1.614/

Dataset Link: https://ltrc.iiit.ac.in/showfile.php?filename=downloads/teSum/

Work from IIIT Hyderabad

[kavyamanohar]

The paper link is: https://aclanthology.org/2022.lrec-1.614/

[anoopkunchukuttan]

Thanks for pointing out - that was a mistake