University of Cambridge > Talks.cam > NLIP Seminar Series > Sentence-level Topic Models

  Browse A-Z

Log in

Googledetails Microsoftdetails

๐Ÿ” Log In

Information on

Subscribing to talksdetails Finding a talkdetails Adding a talkdetails Disseminating talksdetails Help and Documentationdetails

โ„น๏ธ Information

Sentence-level Topic Models

Download to your calendar using vCal

If you have a question about this talk, please contact Anita Verล‘ .

We present two generative models of documents which generate whole sentences from underlying topics. This relaxes the word exchangeability assumption of traditional generative models of documents to sentence exchangeability, and can hence capture inter-word dependencies that LDA misses. Despite the additional model complexity, model training and inference is still feasible using state-of-the-art approximate inference techniques. We show that both our proposed models achieve lower perplexities than a standard LDA topic model and a strong LSTM language model on held-out documents. We also manually inspect samples from the topics learnt, and show that the topics both models learn are coherent. Finally, we show that on a shallow document classification task, LDA outperforms our models, and analyse the reasons behind this.

This talk is part of the NLIP Seminar Series series.

This talk is included in these lists:

Note that ex-directory lists are not shown.

Other lists

Sociology Seminar Series 22-23 Maxwell Lecture Theatre Goldsmith's 2 & https://us06web.zoom.us/j/81688365420?pwd=b3RpdHgxNVhWbWQrbXluMUxYNWVtQT09 Meeting ID: 816 8836 5420 Passcode: 047859

Other talks

Disentangling preventative & ejective AGN feedback on a population basis with simulations, observations & machine learning Talk title TBC In conversation: Louis Herns Marcelin and Ola Osman on legacies of enslavement and contemporary violence in Liberia and Haiti External Seminar - Jenn Brophy TBC Topic Film, TV shows for Medical Education, โ€˜The Wireโ€™ as a case study Alicia Anderson discussing some exciting news about HARPS3
 

ยฉ 2006-2025 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity