Team Alpage

Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

Coreference Resolution

Participant :

Early machine learning approaches to coreference resolution rely on local, discriminative pairwise classifiers [113] , [100] , [98] made considerable progress in creating robust coreference systems, but their performance still left much room for improvement. This stems from two main deficiencies:

More recent work has sought to address these limitations. For example, to address decision locality, McCallum and Wellner [95] use conditional random fields with model structures in which pairwise decisions influence others. Denis [85] and Klenner [90] use integer linear programming (ilp ) to perform global inference via transitivity constraints between different coreference decisions. Denis and Baldridge [84] use a ranker to compare antecedents for an anaphor simultaneously rather than in the standard pairwise manner. To address the knowledge bottleneck problem, Denis and Baldridge [6] use ilp for joint inference using a pairwise coreference model and a model for determining the anaphoricity of mentions. Also, Denis and Baldridge [84] and Bengston and Roth [61] use models and features, respectively, that attend to particular types of mentions (e.g., full noun phrases versus pronouns). Furthermore, Bengston and Roth [61] use a wider range of features than are normally considered, and in particular use predicted features for later classifiers, to considerably boost performance.

In [13] , we use ilp to extend the joint formulation of Denis and Baldridge [6] using named entity classification and combine it with the transitivity constraints [85] , [90] . Intuitively, we only should identify antecedents for the mentions which are likely to have one [99] , and we should only make a set of mentions coreferent if they are all instances of the same entity type (eg, person or location ). ilp enables such constraints to be declared between the outputs of independent classifiers to ensure coherent assignments are made. It also leads to global inference via both constraints on named entity types and transitivity constraints since both relate multiple pairwise decisions.

We show that this strategy leads to improvements across the three main metrics proposed for coreference: the muc metric [116] , the b3 metric [60] , and ceaf metric [92] . In addition, we contextualize the performance of our system with respect to cascades of multiple models and oracle systems that assume perfect information (e.g. about entity types). We furthermore demonstrate the inadequacy of using only the muc metric and argue that results should always be given for all three.


Logo Inria