We describe challenges and advantages unique to coreference resolution in the biomedical domain, and a sieve-based architecture that leverages domain knowledge for both entity and event coreference resolution. Domain-general coreference resolution algorithms perform poorly on biomedical documents, because the cues they rely on such as gender are largely absent in this domain, and because they do not encode domain-specific knowledge such as the number and type of participants required in chemical reactions. Moreover, it is difficult to directly encode this knowledge into most coreference resolution algorithms because they are not rule-based. Our rule-based architecture uses sequentially applied hand-designed "sieves", with the output of each sieve informing and constraining subsequent sieves. This architecture provides a 3.2% increase in throughput to our Reach event extraction system with precision parallel to that of the stricter system that relies solely on syntactic patterns for extraction.
Coreference resolution systems are typically trained with heuristic loss functions that require careful tuning. In this paper we instead apply reinforcement learning to directly optimize a neural mention-ranking model for coreference evaluation metrics. We experiment with two approaches: the REINFORCE policy gradient algorithm and a reward-rescaled max-margin objective. We find the latter to be more effective, resulting in significant improvements over the current state-of-the-art on the English and Chinese portions of the CoNLL 2012 Shared Task.
In this paper, we propose commonsense knowledge enhanced embeddings (KEE) for solving the Pronoun Disambiguation Problems (PDP). The PDP task we investigate in this paper is a complex coreference resolution task which requires the utilization of commonsense knowledge. This task is a standard first round test set in the 2016 Winograd Schema Challenge. In this task, traditional linguistic features that are useful for coreference resolution, e.g. context and gender information, are no longer effective anymore. Therefore, the KEE models are proposed to provide a general framework to make use of commonsense knowledge for solving the PDP problems. Since the PDP task doesn't have training data, the KEE models would be used during the unsupervised feature extraction process. To evaluate the effectiveness of the KEE models, we propose to incorporate various commonsense knowledge bases, including ConceptNet, WordNet, and CauseCom, into the KEE training process. We achieved the best performance by applying the proposed methods to the 2016 Winograd Schema Challenge. In addition, experiments conducted on the standard PDP task indicate that, the proposed KEE models could solve the PDP problems by achieving 66.7% accuracy, which is a new state-of-the-art performance.
Progress in text understanding has been driven by large datasets that test particular capabilities, like recent datasets for reading comprehension (Hermann et al., 2015). We focus here on the LAMBADA dataset (Paperno et al., 2016), a word prediction task requiring broader context than the immediate sentence. We view LAMBADA as a reading comprehension problem and apply comprehension models based on neural networks. Though these models are constrained to choose a word from the context, they improve the state of the art on LAMBADA from 7.3% to 49%. We analyze 100 instances, finding that neural network readers perform well in cases that involve selecting a name from the context based on dialogue or discourse cues but struggle when coreference resolution or external knowledge is needed.
Only a year ago, all state-of-the-art coreference resolvers were using an extensive amount of surface features. Recently, there was a paradigm shift towards using word embeddings and deep neural networks, where the use of surface features is very limited. In this paper, we show that a simple SVM model with surface features outperforms more complex neural models for detecting anaphoric mentions. Our analysis suggests that using generalized representations and surface features have different strength that should be both taken into account for improving coreference resolution.
Lexical features are a major source of information in state-of-the-art coreference resolvers. Lexical features implicitly model some of the linguistic phenomena at a fine granularity level. They are especially useful for representing the context of mentions. In this paper we investigate a drawback of using many lexical features in state-of-the-art coreference resolvers. We show that if coreference resolvers mainly rely on lexical features, they can hardly generalize to unseen domains. Furthermore, we show that the current coreference resolution evaluation is clearly flawed by only evaluating on a specific split of a specific dataset in which there is a notable overlap between the training, development and test sets.
This paper introduces a new model that uses named entity recognition, coreference resolution, and entity linking techniques, to approach the task of linking people entities on Wikipedia people pages to their corresponding Wikipedia pages if applicable. Our task is different from general and traditional entity linking because we are working in a limited domain, namely, people entities, and we are including pronouns as entities, whereas in the past, pronouns were never considered as entities in entity linking. We have built 2 models, both outperforms our baseline model significantly. The purpose of our project is to build a model that could be use to generate cleaner data for future entity linking tasks. Our contribution include a clean data set consisting of 50Wikipedia people pages, and 2 entity linking models, specifically tuned for this domain.
We describe the Marmara Turkish Coreference Corpus, which is an annotation of the whole METU-Sabanci Turkish Treebank with mentions and coreference chains. Collecting eight or more independent annotations for each document allowed for fully automatic adjudication. We provide a baseline system for Turkish mention detection and coreference resolution and evaluate it on the corpus.
Selectional preferences have long been claimed to be essential for coreference resolution. However, they are mainly modeled only implicitly by current coreference resolvers. We propose a dependency-based embedding model of selectional preferences which allows fine-grained compatibility judgments with high coverage. We show that the incorporation of our model improves coreference resolution performance on the CoNLL dataset, matching the state-of-the-art results of a more complex system. However, it comes with a cost that makes it debatable how worthwhile such improvements are.
We introduce a novel iterative approach for event coreference resolution that gradually builds event clusters by exploiting inter-dependencies among event mentions within the same chain as well as across event chains. Among event mentions in the same chain, we distinguish within- and cross-document event coreference links by using two distinct pairwise classifiers, trained separately to capture differences in feature distributions of within- and cross-document event clusters. Our event coreference approach alternates between WD and CD clustering and combines arguments from both event clusters after every merge, continuing till no more merge can be made. And then it performs further merging between event chains that are both closely related to a set of other chains of events. Experiments on the ECB+ corpus show that our model outperforms state-of-the-art methods in joint task of WD and CD event coreference resolution.
Adding manually annotated prosodic information, specifically pitch accents and phrasing, to the typical text-based feature set for coreference resolution has previously been shown to have a positive effect on German data. Practical applications on spoken language, however, would rely on automatically predicted prosodic information. In this paper we predict pitch accents (and phrase boundaries) using a convolutional neural network (CNN) model from acoustic features extracted from the speech signal. After an assessment of the quality of these automatic prosodic annotations, we show that they also significantly improve coreference resolution.
Coreference resolution is an intermediate step for text understanding. It is used in tasks and domains for which we do not necessarily have coreference annotated corpora. Therefore, generalization is of special importance for coreference resolution. However, while recent coreference resolvers have notable improvements on the CoNLL dataset, they struggle to generalize properly to new domains or datasets. In this paper, we investigate the role of linguistic features in building more generalizable coreference resolvers. We show that generalization improves only slightly by merely using a set of additional linguistic features. However, employing features and subsets of their values that are informative for coreference resolution, considerably improves generalization. Thanks to better generalization, our system achieves state-of-the-art results in out-of-domain evaluations, e.g., on WikiCoref, our system, which is trained on CoNLL, achieves on-par performance with a system designed for this dataset.
Identifying nominals with no head match is a long-standing challenge in coreference resolution with current systems performing significantly worse than humans. In this paper we present a new neural network architecture which outperforms the current state-of-the-art system on the English portion of the CoNLL 2012 Shared Task. This is done by using a logistic regression on features produced by two submodels, one of which is has the architecture proposed in [CM16a] while the other combines domain specific embeddings of the antecedent and the mention. We also propose some simple additional features which seem to improve performance for all models substantially, increasing F1 by almost 4% on basic logistic regression and other complex models.
In this paper, we demonstrate the importance of coreference resolution for natural language processing on the example of the TAC Slot Filling shared task. We illustrate the strengths and weaknesses of automatic coreference resolution systems and provide experimental results to show that they improve performance in the slot filling end-to-end setting. Finally, we publish KBPchains, a resource containing automatically extracted coreference chains from the TAC source corpus in order to support other researchers working on this topic.
In this paper, we describe TAMU's system submitted to the TAC KBP 2017 event nugget detection and coreference resolution task. Our system builds on the statistical and empirical observations made on training and development data. We found that modifiers of event nuggets tend to have unique syntactic distribution. Their parts-of-speech tags and dependency relations provides them essential characteristics that are useful in identifying their span and also defining their types and realis status. We further found that the joint modeling of event span detection and realis status identification performs better than the individual models for both tasks. Our simple system designed using minimal features achieved the micro-average F1 scores of 57.72, 44.27 and 42.47 for event span detection, type identification and realis status classification tasks respectively. Also, our system achieved the CoNLL F1 score of 27.20 in event coreference resolution task.
We introduce a new benchmark, WinoBias, for coreference resolution focused on gender bias. Our corpus contains Winograd-schema style sentences with entities corresponding to people referred by their occupation (e.g. the nurse, the doctor, the carpenter). We demonstrate that a rule-based, a feature-rich, and a neural coreference system all link gendered pronouns to pro-stereotypical entities with higher accuracy than anti-stereotypical entities, by an average difference of 21.1 in F1 score. Finally, we demonstrate a data-augmentation approach that, in combination with existing word-embedding debiasing techniques, removes the bias demonstrated by these systems in WinoBias without significantly affecting their performance on existing coreference benchmark datasets. Our dataset and code are available at http://winobias.org.
Notional anaphors are pronouns which disagree with their antecedents' grammatical categories for notional reasons, such as plural to singular agreement in: 'the government ... they'. Since such cases are rare and conflict with evidence from strictly agreeing cases ('the government ... it'), they present a substantial challenge to both coreference resolution and referring expression generation. Using the OntoNotes corpus, this paper takes an ensemble approach to predicting English notional anaphora in context on the basis of the largest empirical data to date. In addition to state of the art prediction accuracy, the results suggest that theoretical approaches positing a plural construal at the antecedent's utterance are insufficient, and that circumstances at the anaphor's utterance location, as well as global factors such as genre, have a strong effect on the choice of referring expression.
Multi-modal data is becoming more common in big data background. Finding the semantically similar objects from different modality is one of the heart problems of multi-modal learning. Most of the current methods try to learn the inter-modal correlation with extrinsic supervised information, while intrinsic structural information of each modality is neglected. The performance of these methods heavily depends on the richness of training samples. However, obtaining the multi-modal training samples is still a labor and cost intensive work. In this paper, we bring a extrinsic correlation between the space structures of each modalities in coreference resolution. With this correlation, a semi-supervised learning model for multi-modal coreference resolution is proposed. We firstly extract high-level features of images and text, then compute the distances of each object from some reference points to build the space structure of each modality. With a shared reference point set, the space structures of each modality are correlated. We employ the correlation to build a commonly shared space that the semantic distance between multi-modal objects can be computed directly. The experiments on two multi-modal datasets show that our model performs better than the existing methods with insufficient training data.
We present an empirical study of gender bias in coreference resolution systems. We first introduce a novel, Winograd schema-style set of minimal pair sentences that differ only by pronoun gender. With these "Winogender schemas," we evaluate and confirm systematic gender bias in three publicly-available coreference resolution systems, and correlate this bias with real-world and textual gender statistics.
The package cleanNLP provides a set of fast tools for converting a textual corpus into a set of normalized tables. The underlying natural language processing pipeline utilizes Stanford's CoreNLP library, exposing a number of annotation tasks for text written in English, French, German, and Spanish. Annotators include tokenization, part of speech tagging, named entity recognition, entity linking, sentiment analysis, dependency parsing, coreference resolution, and information extraction.