We describe a set of diverse podcast information needs and different approaches to assessing retrieved content for relevance.
no code implementations • 17 Jun 2021 • Rosie Jones, Hamed Zamani, Markus Schedl, Ching-Wei Chen, Sravana Reddy, Ann Clifton, Jussi Karlgren, Helia Hashemi, Aasish Pappu, Zahra Nazari, Longqi Yang, Oguz Semerci, Hugues Bouchard, Ben Carterette
Podcasts are spoken documents across a wide-range of genres and styles, with growing listenership across the world, and a rapidly lowering barrier to entry for both listeners and creators.
This paper contains the description of our submissions to the summarization task of the Podcast Track in TREC (the Text REtrieval Conference) 2020.
The Podcast Track is new at the Text Retrieval Conference (TREC) in 2020.
Paired with the audio files, they are also a resource for speech processing and the study of paralinguistic, sociolinguistic, and acoustic aspects of the domain.
Written in Python and built on MXNet, the toolkit offers scalable training and inference for the three most prominent encoder-decoder architectures: attentional recurrent neural networks, self-attentional transformers, and fully convolutional networks.
We apply stochastic average gradient (SAG) algorithms for training conditional random fields (CRFs).