Global inference for sentence compression : an integer linear programming approach

By James Clarke, Mirella Lapata. Journal of Artificial Intelligence Research, 2008

331 citations recorded by the source corpus.

Sentence compression holds promise for many applications ranging from summarization to subtitle generation. Our work views sentence compression as an optimization problem and uses integer linear programming (ILP) to infer globally optimal compressions in the presence of linguistically motivated constraints. We show how previous formulations of sentence compression can be recast as ILPs and extend these models with novel global constraints. Experimental results on written and spoken texts demonstrate improvements over state-of-the-art models.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/1ec86811a79fb02a1c551b8f418314a00f5f5a99

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

UvA-DARE (Digital Academic Repository) A Dataset and Evaluation Metrics for Abstractive Compression of Sentences and Short Paragraphs (unknown): “Prior studies of human compression: Clarke (2008) studied the properties of manually-collected deletion-based compressions in the news genre, comparing them with automatically-mined data from the Ziff-Davis corpus in terms of compression rate, length of deleted spans, and deletion probability by…”
From Extractive to Abstractive Summarization: A Journey (2019, methodology): “[10] treat sentence compression as an integer linear programming problem.”
Multi-Reward Reinforced Summarization with Saliency and Entailment (2018, methodology): “Earlier summarization work was based on extraction and compression-based approaches (Jing, 2000; Knight and Marcu, 2002; Clarke and Lapata, 2008; Filippova et al., 2015), with more focus on graph-based (Giannakopoulos, 2009; Ganesan et al.”
Jointly Learning to Extract and Compress (2011, background): “While pure extraction is certainly simple and does guarantee some minimal readability, Lin (2003) showed that sentence compression (Knight and Marcu, 2001; McDonald, 2006; Clarke and Lapata, 2008) has the potential to improve the resulting summaries.”
TreeTalk: Composition and Compression of Trees for Image Descriptions (2014, methodology): “Therefore, we encode our problem using Integer Linear Programming (ILP) (Roth and tau Yih, 2004; Clarke and Lapata, 2008) and use the CPLEX (ILOG, Inc, 2006) solver.”
Retrieve, Rerank and Rewrite: Soft Template Based Neural Summarization (2018, methodology): “tion include template-based methods (Zhou and Hovy, 2004), syntactic tree pruning (Knight and Marcu, 2002; Clarke and Lapata, 2008) and statistical machine translation techniques (Banko et al.”
Soft Layer-Specific Multi-Task Summarization with Entailment and Question Generation (2018, unknown): “Automatic text summarization has been progressively improving over the time, initially more focused on extractive and compressive models (Jing and McKeown, 2000; Knight and Marcu, 2002; Clarke and Lapata, 2008; Filippova et al., 2015; Kedzie et al., 2015), and moving more towards compressive and abs”
Joint Inference for Event Timeline Construction (2012, methodology): “We perform exact inference using Integer Linear Programming (ILP) as in (Roth and Yih, 2007; Clarke and Lapata, 2008).”
A Neural Attention Model for Abstractive Sentence Summarization (2015, background): “The ﬁrst is a sentence compression baseline C OM - PRESS (Clarke and Lapata, 2008).”
A Survey of Paraphrasing and Textual Entailment Methods (2009, background): “Written News Compression Corpus: Sentence compression corpus; http://jamesclarke.net/research/.”
Domain-Specific Image Captioning (2014, background): “Sequential constraints are deﬁned as in Clarke (2008) ensure that the ordering of the tri-grams is valid, and that the mathematical validity of the model holds.”
Automatic Generation of Story Highlights (2010, background): “Despite the bulk of work on sentence compression and summarization (see Clarke and Lapata 2008 and Mani 2001 for overviews) only a handful of approaches attempt to do both in a joint model (Daum´e III and Marcu, 2002; Daum´e III, 2006; Lin, 2003; Martins and Smith, 2009).”
Discourse Constraints for Document Compression (2010, background): “All three systems could use further improvement, as inter-annotator agreement on this data yields an F1 of 65.8% (Clarke 2008).”
Phrase-based Compressive Cross-Language Summarization (2015, methodology): “The compression process follows from an integer linear program as described by Clarke and Lapata (2008). This baseline gives strong performance as we have found on English DUC 2001 dataset as well as other monolingual datasets.”
Joint Parsing and Generation for Abstractive Summarization (2019, background): “A number of recent studies have explored constituency and dependency grammars (Daum´e III and Marcu 2002; Clarke and Lapata 2008; Martins and Smith 2009; Filippova 2010; Berg-Kirkpatrick, Gillick, and Klein 2011; Wang et al. 2013; Durrett, Berg-Kirkpatrick, and Klein 2016), rhetorical structure…”
Recent advances in document summarization (2017, background): “as well as suitable units in order to make automatic metrics more reliable, while neural models currently are still inferior to classic deletion-based ILP frameworks [34] in terms of human ratings.”
Learning-Based Single-Document Summarization with Compression and Anaphoricity Constraints (2016, methodology): “Following work on isolated sentence compression (McDonald, 2006; Clarke and Lapata, 2008) and”