We present a corpus of anaphoric information (coreference) crowdsourced through a gamewith-a-purpose. The corpus, containing annotations for about 108,000 markables, is one of the largest corpora for coreference for English, and one of the largest crowdsourced NLP corpora, but its main feature is the large number of judgments per markable: 20 on average, and over 2.2M in total. This characteristic makes the corpus a unique resource for the study of disagreements on anaphoric interpretation. A second distinctive feature is its rich annotation scheme, covering singletons, expletives, and split-antecedent plurals. Finally, the corpus also comes with labels inferred using a recently proposed probabilistic model of annotation for coreference. The labels are of high quality and make it possible to successfully train a state of the art coreference resolver, including training on singletons and non-referring expressions. The annotation model can also result in more than one label, or no label, being proposed for a markable, thus serving as a baseline method for automatically identifying ambiguous markables. A preliminary analysis of the results is presented.

Many tasks in Natural Language Processing (NLP) and Computer Vision (CV) offer evidence that humans disagree, from objective tasks such as part-of-speech tagging to more subjective tasks such as classifying an image or deciding whether a proposition follows from certain premises. While most learning in artificial intelligence (AI) still relies on the assumption that a single (gold) interpretation exists for each item, a growing body of research aims to develop learning methods that do not rely on this assumption. In this survey, we review the evidence for disagreements on NLP and CV tasks, focusing on tasks for which substantial datasets containing this information have been created. We discuss the most popular approaches to training models from datasets containing multiple judgments potentially in disagreement. We systematically compare these different approaches by training them with each of the available datasets, considering several ways to evaluate the resulting models. Finally, we discuss the results in depth, focusing on four key research questions, and assess how the type of evaluation and the characteristics of a dataset determine the answers to these questions. Our results suggest, first of all, that even if we abandon the assumption of a gold standard, it is still essential to reach a consensus on how to evaluate models. This is because the relative performance of the various training methods is critically affected by the chosen form of evaluation. Secondly, we observed a strong dataset effect. With substantial datasets, providing many judgments by high-quality coders for each item, training directly with soft labels achieved better results than training from aggregated or even gold labels. This result holds for both hard and soft evaluation. But when the above conditions do not hold, leveraging both gold and soft labels generally achieved the best results in the hard evaluation. All datasets and models employed in this paper are freely available as supplementary materials.

The analysis of crowdsourced annotations in natural language processing is concerned with identifying (1) gold standard labels, (2) annotator accuracies and biases, and (3) item difficulties and error patterns. Traditionally, majority voting was used for 1, and coefficients of agreement for 2 and 3. Lately, model-based analysis of corpus annotations have proven better at all three tasks. But there has been relatively little work comparing them on the same datasets. This paper aims to fill this gap by analyzing six models of annotation, covering different approaches to annotator ability, item difficulty, and parameter pooling (tying) across annotators and items. We evaluate these models along four aspects: comparison to gold labels, predictive accuracy for new annotations, annotator characterization, and item difficulty, using four datasets with varying degrees of noise in the form of random (spammy) annotators. We conclude with guidelines for model selection, application, and implementation.

Supervised learning assumes that a ground truth label exists. However, the reliability of this ground truth depends on human annotators, who often disagree. Prior work has shown that this disagreement can be helpful in training models. We propose a novel method to incorporate this disagreement as information: in addition to the standard error computation, we use soft labels (i.e., probability distributions over the annotator labels) as an auxiliary task in a multi-task neural network. We measure the divergence between the predictions and the target soft labels with several loss-functions and evaluate the models on various NLP tasks. We find that the soft-label prediction auxiliary task reduces the penalty for errors on ambiguous entities and thereby mitigates overfitting. It significantly improves performance across tasks beyond the standard approach and prior work.

Now that the performance of coreference resolvers on the simpler forms of anaphoric reference has greatly improved, more attention is devoted to more complex aspects of anaphora. One limitation of virtually all coreference resolution models is the focus on single-antecedent anaphors. Plural anaphors with multiple antecedents-so-called split-antecedent anaphors (as in John met Mary. They went to the movies)-have not been widely studied, because they are not annotated in ONTONOTES and are relatively infrequent in other corpora. In this paper, we introduce the first model for unrestricted resolution of split-antecedent anaphors. We start with a strong baseline enhanced by BERT embeddings, and show that we can substantially improve its performance by addressing the sparsity issue. To do this, we experiment with auxiliary corpora where split-antecedent anaphors were annotated by the crowd, and with transfer learning models using element-of bridging references and single-antecedent coreference as auxiliary tasks. Evaluation on the gold annotated ARRAU corpus shows that the out best model uses a combination of three auxiliary corpora achieved F1 scores of 70% and 43.6% when evaluated in a lenient and strict setting, respectively, i.e., 11 and 21 percentage points gain when compared with our baseline. 1(1)The Joneses i went to the park. They i had a good time.(2) John and Mary i went to the park. They i had a good time.(3) John i met Mary j in the park. They i,j had a good chat .(4) John likes green i , Mary likes blue j , but Tom likes both colours i,j .