2018

Summarizing Opinions: Aspect Extraction Meets Sentiment Prediction and They Are Both Weakly Supervised

Angelidis, Stefanos, Lapata, Mirella

Understand

We present a neural framework for opinion summarization from online product reviews which is knowledge-lean and only requires light supervision (e.g., in the form of product domain labels and user-provided ratings).

Our method combines two weakly supervised components to identify salient opinions and form extractive summaries from multiple reviews: an aspect extractor trained under a multi-task objective, and a sentiment predictor based on multiple instance learning.
We introduce an opinion summarization dataset that includes a training set of product reviews from six diverse domains and human-annotated development and test sets with gold standard aspect annotations, salience labels, and opinion summaries.
Automatic evaluation shows significant improvements over baselines, and a large-scale study indicates that our opinion summaries are preferred by human judges according to multiple criteria.

Built on

Rhetorical structure theory: Toward a functional theory of text organization
William C. Mann and Sandra A. Thompson. 1988 · 1988
Earlier work this paper cites.
Best-worst scaling: A model for the largest difference judgments
Jordan J Louviere and George G Woodworth. 1991 · 1991
Earlier work this paper cites.
Predicting query performance
Steve Cronen-Townsend, Yun Zhou, and W. Bruce Croft. 2002 · 2002
Earlier work this paper cites.
Automatic evaluation of summaries using n-gram co-occurrence statistics
Chin-Yew Lin and Eduard Hovy. 2003 · 2003
Earlier work this paper cites.
Evaluation challenges in large-scale document summarization
Dragomir R. Radev, Simone Teufel, Horacio Saggion, Wai Lam, John Blitzer, Hong Qi, Arda Çelebi, Danyu Liu, and Elliott Drabek. 2003 · 2003
Earlier work this paper cites.
Lexrank: Graph-based lexical centrality as salience in text summarization
Günes Erkan and Dragomir R Radev. 2004 · 2004
Earlier work this paper cites.
Mining and summarizing customer reviews
Minqing Hu and Bing Liu. 2004 · 2004
Earlier work this paper cites.
Statistical significance tests for machine translation evaluation
Philipp Koehn. 2004 · 2004
Earlier work this paper cites.
The impact of frequency on summarization
Ani Nenkova and Lucy Vanderwende. 2005 · 2005
Earlier work this paper cites.
Extracting product features and opinions from reviews
Ana-Maria Popescu and Oren Etzioni. 2005 · 2005
Earlier work this paper cites.
Hierarchical clustering algorithms for document datasets
Ying Zhao, George Karypis, and Usama Fayyad. 2005 · 2005
Earlier work this paper cites.
Opinion extraction and summarization on the web
Minqing Hu and Bing Liu. 2006 · 2006
Earlier work this paper cites.
Opinion extraction, summarization and tracking in news and blog corpora
Lun-Wei Ku, Yun-Ting Liang, and Hsin-Hsi Chen. 2006 · 2006
Earlier work this paper cites.
Topic sentiment mixture: Modeling facets and opinions in weblogs
Qiaozhu Mei, Xu Ling, Matthew Wondra, Hang Su, and ChengXiang Zhai. 2007 · 2007
Earlier work this paper cites.
Building a sentiment summarizer for local service reviews
Sasha Blair-Goldensohn, Kerry Hannan, Ryan McDonald, Tyler Neylon, George Reis, and Jeff Reynar. 2008 · 2008
Earlier work this paper cites.

Similar

Sentiment summarization: Evaluating and learning user preferences
Kevin Lerman, Sasha Blair-Goldensohn, and Ryan McDonald. 2009 · 2009
Cited alongside, same era.
Rated aspect summarization of short comments
Yue Lu, ChengXiang Zhai, and Neel Sundaresan. 2009 · 2009
Cited alongside, same era.
Maxdiff analysis: Simple counting, individual-level logit, and HB
Bryan Orme. 2009 · 2009
Cited alongside, same era.
Opinosis: a graph-based approach to abstractive summarization of highly redundant opinions
Kavita Ganesan, ChengXiang Zhai, and Jiawei Han. 2010 · 2010
Cited alongside, same era.
Comprehensive review of opinion summarization
Hyun Duk Kim, Kavita Ganesan, Parikshit Sondhi, and ChengXiang Zhai. 2011 · 2011
Cited alongside, same era.
Text-level discourse parsing with rich linguistic features
Wei Vanessa Feng and Graeme Hirst. 2012 · 2012
Cited alongside, same era.
Sentiment analysis and opinion mining
Bing Liu. 2012 · 2012
Cited alongside, same era.
A survey of opinion mining and sentiment analysis
Bing Liu and Lei Zhang. 2012 · 2012
Cited alongside, same era.
Modeling review comments
Arjun Mukherjee and Bing Liu. 2012 · 2012
Cited alongside, same era.
Distributed representations of words and phrases and their compositionality
Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013 · 2013
Cited alongside, same era.
Recursive deep models for semantic compositionality over a sentiment treebank
Richard Socher, Alex Perelygin, Jean Wu, Jason Chuang, Christopher D. Manning, Andrew Y. Ng, and Christopher Potts. 2013 · 2013
Cited alongside, same era.
Convolutional neural networks for sentence classification
Yoon Kim. 2014 · 2014
Cited alongside, same era.

Then

From group to individual labels using deep features
Dimitrios Kotzias, Misha Denil, Nando De Freitas, and Padhraic Smyth. 2015 · 2015
Later among the works it cites.
Fine-grained opinion mining with recurrent neural networks and word embeddings
Pengfei Liu, Shafiq Joty, and Helen Meng. 2015 · 2015
Later among the works it cites.
Best-worst scaling: Theory, methods and applications
Jordan J Louviere, Terry N Flynn, and Anthony Alfred John Marley. 2015 · 2015
Later among the works it cites.
Image-based recommendations on styles and substitutes
Julian McAuley, Christopher Targett, Qinfeng Shi, and Anton Van Den Hengel. 2015 · 2015
Later among the works it cites.
Feuding families and former friends: Unsupervised learning for dynamic fictional relationships
Mohit Iyyer, Anupam Guha, Snigdha Chaturvedi, Jordan Boyd-Graber, and Hal Daumé III. 2016 · 2016
Later among the works it cites.
The role of discourse units in near-extractive summarization
Junyi Jessy Li, Kapil Thadani, and Amanda Stent. 2016 · 2016
Later among the works it cites.
Neural network-based abstract generation for opinions and arguments
Lu Wang and Wang Ling. 2016 · 2016
Later among the works it cites.
Recursive neural conditional random fields for aspect-based sentiment analysis
Wenya Wang, Sinno Jialin Pan, Daniel Dahlmeier, and Xiaokui Xiao. 2016 · 2016
Later among the works it cites.
Unsupervised word and dependency path embeddings for aspect term extraction
Yichun Yin, Furu Wei, Li Dong, Kaimeng Xu, Ming Zhang, and Ming Zhou. 2016 · 2016
Later among the works it cites.
Improving multi-document summarization via text classification
Ziqiang Cao, Wenjie Li, Sujian Li, and Furu Wei. 2017 · 2017
Later among the works it cites.
An unsupervised neural attention model for aspect extraction
Ruidan He, Wee Sun Lee, Hwee Tou Ng, and Daniel Dahlmeier. 2017 · 2017
Later among the works it cites.
Best-worst scaling more reliable than rating scales: A case study on sentiment intensity annotation
Svetlana Kiritchenko and Saif Mohammad. 2017 · 2017
Later among the works it cites.
Explicit document modeling through weighted multiple-instance learning
Nikolaos Pappas and Andrei Popescu-Belis. 2017 · 2017
Later among the works it cites.
Graph-based neural multi-document summarization
Michihiro Yasunaga, Rui Zhang, Kshitijh Meelu, Ayush Pareek, Krishnan Srinivasan, and Dragomir Radev. 2017 · 2017
Later among the works it cites.
Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis
Stefanos Angelidis and Mirella Lapata. 2018 · 2018
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…

Reading the bibliography…