A Constrained Latent Variable Model for Coreference Resolution

By Kai-Wei Chang, Rajhans Samdani, Dan Roth. 2013

74 citations recorded by the source corpus.

Coreference resolution is a well known clustering task in Natural Language Processing. In this paper, we describe the Latent Left Linking model (L 3 M), a novel, principled, and linguistically motivated latent structured prediction approach to coreference resolution. We show that L 3 M admits efficient inference and can be augmented with knowledge-based constraints; we also present a fast stochastic gradient based learning. Experiments on ACE and Ontonotes data show that L 3 M and its constrained version, CL 3 M, are more accurate than several state-of-the-art approaches as well as some structured prediction models proposed in the literature.

Subfields: Artificial Intelligence, Natural Language Processing

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 Natural Language Processing

Citations in the corpus

Reasoning about Quantities in Natural Language (2015, methodology): “For our classifiers, we use a sparse averaged perceptron implemented with the SNOW framework (Carlson et al., 1999).”
Solving Hard Coreference Problems (2019, methodology): “Baseline Systems: We use the state-of-art Illinois coreference system as our baseline system (Chang et al., 2013).”
The Gap of Semantic Parsing: A Survey on Automatic Math Word Problem Solvers (2018, methodology): “As to applying coreference resolvers in MWP sovers, the Illinois Coreference Resolver [74] [76] is used in [20] to identify pronoun referents and facilitate semantic labeling.”
A Primal Dual Formulation For Deep Learning With Constraints (2019, methodology): “Constrained conditional models use integer linear programming to perform inference with global hard constraints (Roth and Yih [2005], Chang et al. [2013]).”
Entity-Centric Coreference Resolution with Model Stacking (2015, methodology): “work entirely or almost entirely over pairs of mentions (Fernandes et al., 2012; Durrett and Klein, 2013; Chang et al., 2013).”
Learning Structured Perceptrons for Coreference Resolution with Latent Antecedents and Non-local Features (2014, methodology): “Chang et al. (2013) use a max-margin approach to learn a pairwise model and rely on stochastic gradient descent to circumvent the costly operation of decoding the entire training set in order to compute the gradients and the latent antecedents.”
Learning from Explicit and Implicit Supervision Jointly For Algebra Word Problems (2016, background): “Similar efforts have been made to develop latent structured prediction models (Yu and Joachims, 2009; Chang et al., 2013; Zettlemoyer and Collins, 2007) to ﬁnd latent semantic structures which best explain the answer given the question.”
Learning Global Features for Coreference Resolution (2016, methodology): “, 2001; Ng and Cardie, 2002; Bengtson and Roth, 2008), and mention-ranking models, which select a single antecedent for each anaphoric mention (Denis and Baldridge, 2008; Rahman and Ng, 2009; Durrett and Klein, 2013; Chang et al., 2013; Wiseman et al., 2015).”
Robustness in Coreference Resolution (2020, methodology): “• learning from best antecedents during training, selecting best antecedents automatically: The ranking models used by Chang et al. (2012), Wiseman et al. (2015), and Clark & Manning (2016a) are examples of such methods.”
A Joint Model for Entity Analysis: Coreference, Typing, and Linking (2014, methodology): “The systems of Chang et al. (2013) and Webster and Curran (2014) perform similarly to the FERNANDES system; changes in the reference implementation of the metrics make exact comparison to printed numbers difficult.”
A Semantic Loss Function for Deep Learning with Symbolic Knowledge (2017, methodology): “For example, constrained conditional models extend linear models with constraints that are enforced through integer linear programming (Chang et al., 2008; 2013).”
Pre-training of Mention Representations in Coreference Models (2020, background): “Almost all such models rely on vector representations of mentions (Clark and Manning, 2016; Lee et al., 2017, 2018; Denis and Baldridge, 2008; Rahman and Ng, 2009; Durrett et al., 2013; Chang et al., 2013; Wiseman et al., 2016; Martschat and Strube, 2015).”