Simple Coreference Resolution with Rich Syntactic and Semantic Features

By Aria Haghighi, Daniel Klein. 2009

231 citations recorded by the source corpus.

Coreference systems are driven by syntactic, semantic, and discourse constraints. We present a simple approach which completely modularizes these three aspects. In contrast to much current work, which focuses on learning and on the discourse component, our system is deterministic and is driven entirely by syntactic and semantic compatibility as learned from a large, unlabeled corpus. Despite its simplicity and discourse naivete, our system substantially outperforms all unsupervised systems and most supervised ones. Primary contributions include (1) the presentation of a simple-to-reproduce, high-performing baseline and (2) the demonstration that most remaining errors can be attributed to syntactic and semantic factors external to the coreference phenomenon (and perhaps best addressed by non-coreference systems).

Subfields: Artificial Intelligence, Information Extraction

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 Information Extraction

Citations in the corpus

Easy Victories and Uphill Battles in Coreference Resolution (2013, background): “Haghighi and Klein, 2010; Rahman and Ng, 2011b) or rule-based (Haghighi and Klein, 2009; Lee et al., 2011), draw on diverse information sources and complex heuristics to resolve pronouns, model discourse, determine anaphoricity, and identify seman-”
Dynamic Knowledge-Base Alignment for Coreference Resolution (2013, background): “A variety of approaches incorporate linguistic knowledge as rules iteratively applied to identify the chains, such as Haghighi and Klein (2009), Raghunathan et al. (2010), Stoyanov et al. (2010). Alternatively (and similar to our approach), others represent this knowledge as features in a machine le”
Biperpedia: An Ontology for Search Applications (2014, methodology): “We run a standard coreference resolution algorithm [13] to ﬁnd if the strings “barack obama” and “president” co-refer to the same entity sufﬁcient number of times in the text corpus.”
A Pronoun Anaphora Resolution System based on Factorial Hidden Markov Models (2011, background): “This was what (Haghighi and Klein, 2009) did and we did this in training with the REUTERS corpus (Hasler et al., 2006) in which syntactic roles are annotated.”
Coreference Resolution with World Knowledge (2011, background): “Previous work has shown that syntactic appositions, which can be extracted using heuristics from unannotated documents or parse trees, are a useful source of world knowledge for coreference resolution (e.g., Daum´e III and Marcu (2005), Ng (2007), Haghighi and Klein (2009)).”
Bootstrapping coreference resolution using word associations (2011, unknown): “Recent work by Haghighi and Klein (2009), Klen-ner and Ailloud (2009) and Raghunathan et al. (2010) challenges the appropriateness of machine learning methods for CoRe.”
A Joint Model for Entity Analysis: Coreference, Typing, and Linking (2014, methodology): “…(2007; Ratinov and Roth (2012) • Infobox type (e.g. Person , Company ) • Copula in the ﬁrst sentence ( is a British politician ); used for coreference previously in Haghighi and Klein (2009) We ﬁre features that conjoin the information from the selected Wikipedia article with the selected NER type.”
Learning-based Multi-Sieve Co-reference Resolution with Knowledge (2012, methodology): “However, rule-based systems, e.g. (Haghighi and Klein, 2009; Raghu-nathan et al., 2010) characterize mention pairs by discourse structure and linguistic properties and apply rules in a prescribed order (high-precision rules ﬁrst).”
Understandable Big Data: A survey (2015, background): “Haghighi and Klein [56] introduce new syntactic, semantic features and discourse phenomena to improve existing systems.”
Resolving Complex Cases of Definite Pronouns: The Winograd Schema Challenge (2012, background): “One reason could be that these difﬁcult pronouns do not appear frequently in standard evaluation corpora such as MUC, ACE, and OntoNotes (Bagga, 1998; Haghighi and Klein, 2009).”
ReNoun: Fact Extraction for Nominal Attributes (2014, methodology): “We rely on a coreference resolver in the spirit of Haghighi and Klein (2009). The resolver clusters the mentions of entities in a document so the references in each cluster are assumed to refer to the same real-world entity.”
Supervised Noun Phrase Coreference Research: The First Fifteen Years (2010, background): “For instance, given the pattern X is a Y (which is highly indicative that X and Y are coreferent), we can instantiate it with a pair of NPs and search for the instantiated pattern in a large corpus or the Web (Daum´e III and Marcu, 2005; Haghighi and Klein, 2009).”
Design and realization of a modular architecture for textual entailment (2013, methodology): “The English LAP includes a pipeline based on the Stanford NLP tools (Toutanova and Manning 2000; Finkel, Grenager and Manning 2005), an easy-first parser (Goldberg and Elhadad 2010), and the Ark coreference resolver (Haghighi and Klein 2009).”
Reading The Web with Learned Syntactic-Semantic Inference Rules (2012, methodology): “We then use a within-document coreference resolver comparable to that of Haghighi and Klein (2009) to group referring NPs into co-referring clusters.”
Joint Coreference Resolution and Named-Entity Linking with Multi-Pass Sieves (2013, methodology): “As the table shows, with gold mentions our system outperforms Haghighi and Klein (2009), Poon and Domingos (2008), Finkel and Man-ning (2008) and the Stanford sieve algorithm across all metrics.”
The Gap of Semantic Parsing: A Survey on Automatic Math Word Problem Solvers (2018, background): “On top of [71], Raghunathan et al. [72] proposed an architecture based on tiers of deterministic coreference models.”
Look Again at the Syntax: Relational Graph Convolutional Network for Gendered Ambiguous Pronoun Resolution (2019, methodology): “Most heuristics-based methods use syntax-based rules (Hobbs, 1978; Lappin and Leass, 1994; Haghighi and Klein, 2009).”
A Multi-Pass Sieve for Coreference Resolution (2010, methodology): “Most coreference resolution approaches perform the task by aggregating local decisions about pairs of mentions (Bengston and Roth, 2008; Finkel and Manning, 2008; Haghighi and Klein, 2009; Stoyanov, 2010).”