Unsupervised Abstractive Opinion Summarization by Generating Sentences with Tree-Structured Topic Guidance



Isonuma, Masaru, Mori, Junichiro, Bollegala, Danushka and Sakata, Ichiro
Unsupervised Abstractive Opinion Summarization by Generating Sentences with Tree-Structured Topic Guidance.

[img] Text
TACL2020.pdf - Accepted Version

Download (690kB) | Preview

Abstract

This paper presents a novel unsupervised abstractive summarization method for opinionated texts. While the basic variational autoencoder-based models assume a unimodal Gaussian prior for the latent code of sentences, we alternate it with a recursive Gaussian mixture, where each mixture component corresponds to the latent code of a topic sentence and is mixed by a tree-structured topic distribution. By decoding each Gaussian component, we generate sentences with tree-structured topic guidance, where the root sentence conveys generic content, and the leaf sentences describe specific topics. Experimental results demonstrate that the generated topic sentences are appropriate as a summary of opinionated texts, which are more informative and cover more input contents than those generated by the recent unsupervised summarization model (Bra\v{z}inskas et al., 2020). Furthermore, we demonstrate that the variance of latent Gaussians represents the granularity of sentences, analogous to Gaussian word embedding (Vilnis and McCallum, 2015).

Item Type: Article
Additional Information: accepted to TACL, pre-MIT Press publication version
Uncontrolled Keywords: cs.CL, cs.CL, cs.AI, cs.LG
Divisions: Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User: Symplectic Admin
Date Deposited: 28 Jun 2021 07:58
Last Modified: 19 Jul 2021 11:36
Related URLs:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3127707