Accepted author manuscript, 214 KB, PDF document
Available under license: CC BY: Creative Commons Attribution 4.0 International License
Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review
Research output: Contribution to conference - Without ISBN/ISSN › Conference paper › peer-review
}
TY - CONF
T1 - Exploring Abstractive Text Summarisation for Podcasts
T2 - 14th Conference on Recent Advances in Natural Language Processing
AU - Saxena, Parth
AU - El-Haj, Mahmoud
PY - 2023/9/6
Y1 - 2023/9/6
N2 - Podcasts have become increasingly popular in recent years, resulting in a massive amount of audio content being produced every day. Efficient summarisation of podcast episodes can enable better content management and discovery for users. In this paper, we explore the use of abstractive text summarisation methods to generate high-quality summaries of podcast episodes. We use pre-trained models, BART and T5, to fine-tune on a dataset of Spotify's 100K podcast. We evaluate our models using automated metrics and human evaluation, and find that the BART model fine-tuned on the podcast dataset achieved a higher ROUGE-1 and ROUGE-L score compared to other models, while the T5 model performed better in terms of semantic meaning. The human evaluation indicates that both models produced high-quality summaries that were well received by participants. Our study demonstrates the effectiveness of abstractive summarisation methods for podcast episodes and offers insights for improving the summarisation of audio content.
AB - Podcasts have become increasingly popular in recent years, resulting in a massive amount of audio content being produced every day. Efficient summarisation of podcast episodes can enable better content management and discovery for users. In this paper, we explore the use of abstractive text summarisation methods to generate high-quality summaries of podcast episodes. We use pre-trained models, BART and T5, to fine-tune on a dataset of Spotify's 100K podcast. We evaluate our models using automated metrics and human evaluation, and find that the BART model fine-tuned on the podcast dataset achieved a higher ROUGE-1 and ROUGE-L score compared to other models, while the T5 model performed better in terms of semantic meaning. The human evaluation indicates that both models produced high-quality summaries that were well received by participants. Our study demonstrates the effectiveness of abstractive summarisation methods for podcast episodes and offers insights for improving the summarisation of audio content.
M3 - Conference paper
Y2 - 4 September 2023 through 6 September 2023
ER -