University of Cambridge > > NLIP Seminar Series > Topic-Aware Convolutional Neural Networks for Extreme Summarization

Topic-Aware Convolutional Neural Networks for Extreme Summarization

Add to your list(s) Download to your calendar using vCal

If you have a question about this talk, please contact Andrew Caines.

We introduce extreme summarization, a new single-document summarization task which does not favor extractive strategies and calls for an abstractive modeling approach. The idea is to create a short, one-sentence news summary answering the question “What is the article about?”. We have collected a real-world, large-scale dataset for this task by harvesting online articles from the British Broadcasting Corporation (BBC). We propose a novel abstractive model which is conditioned on the article’s topics and based entirely on convolutional neural networks. We demonstrate experimentally that this architecture captures long-range dependencies in a document and recognizes pertinent content, outperforming an oracle extractive system and state-of-the-art abstractive approaches when evaluated automatically and by humans.

Bio: Shashi Narayan is a research associate at the School of Informatics at the University of Edinburgh. He will join Google London in February. His research focuses on natural language generation, understanding and structure predictions. A major aim of his research is to build on the hypothesis that tailoring a model with knowledge of the task structure and linguistic requirements, such as syntax and semantics, leads to a better performance. His research has appeared in computational linguistics journals (e.g., TACL and Computational Linguistics) and in conferences (e.g., ACL , EMNLP, NAACL , COLING, EACL and INLG ). He was nominated on the SIGGEN board (2012-14) as a student member. He served as an area co-chair for Generation and gave a tutorial on “Deep learning approaches to text production” at NAACL HLT 2018 . He is currently writing a book on the same topic. He will serve as an area co-chair for Summarization at ACL 2019 .

This talk is part of the NLIP Seminar Series series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.


© 2006-2024, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity