尽管深度神经网络能够在各种任务上实现优于人类的表现，但他们臭名昭著，因为他们需要大量的数据和计算资源，将其成功限制在可用的这些资源的领域。金属学习方法可以通过从相关任务中转移知识来解决此问题，从而减少学习新任务所需的数据和计算资源的数量。我们组织了元数据竞赛系列，该系列为世界各地的研究小组提供了创建和实验评估实际问题的新元学习解决方案的机会。在本文中，我们在竞争组织者和排名最高的参与者之间进行了合作，我们描述了竞争的设计，数据集，最佳实验结果以及Neurips 2021挑战中最高的方法，这些方法吸引了15进入最后阶段的活跃团队（通过表现优于基线），在反馈阶段进行了100多次代码提交。顶级参与者的解决方案是开源的。汲取的经验教训包括学习良好的表示对于有效的转移学习至关重要。

Since a lexicon-based approach is more elegant scientifically, explaining the solution components and being easier to generalize to other applications, this paper provides a new approach for offensive language and hate speech detection on social media. Our approach embodies a lexicon of implicit and explicit offensive and swearing expressions annotated with contextual information. Due to the severity of the social media abusive comments in Brazil, and the lack of research in Portuguese, Brazilian Portuguese is the language used to validate the models. Nevertheless, our method may be applied to any other language. The conducted experiments show the effectiveness of the proposed approach, outperforming the current baseline methods for the Portuguese language.

经典信息提取技术包括建立有关事实的问题和答案。确实，主观信息提取系统在上下文中识别观点和感觉仍然是一个挑战。在基于情感的NLP任务中，在上下文中，最重要的是进攻或仇恨的意见，几乎没有资源来提取信息。为了填补这一重要差距，这篇简短的论文提供了一种新的跨语义和上下文进攻词典，该词典由明确的和隐性的进攻和宣誓的意见表达组成，在两个不同的类别中注释了两种不同的类别：依赖于上下文和与上下文无关的进攻。此外，我们还提供标记来识别仇恨言论。在表达级别评估注释方法，并达到了高通道的一致性。提供的进攻词典有葡萄牙语和英语语言。

Due to the severity of the social media offensive and hateful comments in Brazil, and the lack of research in Portuguese, this paper provides the first large-scale expert annotated corpus of Brazilian Instagram comments for hate speech and offensive language detection. The HateBR corpus was collected from the comment section of Brazilian politicians' accounts on Instagram and manually annotated by specialists, reaching a high inter-annotator agreement. The corpus consists of 7,000 documents annotated according to three different layers: a binary classification (offensive versus non-offensive comments), offensiveness-level classification (highly, moderately, and slightly offensive), and nine hate speech groups (xenophobia, racism, homophobia, sexism, religious intolerance, partyism, apology for the dictatorship, antisemitism, and fatphobia). We also implemented baseline experiments for offensive language and hate speech detection and compared them with a literature baseline. Results show that the baseline experiments on our corpus outperform the current state-of-the-art for the Portuguese language.

Selecting an effective training signal for tasks in natural language processing is difficult: collecting expert annotations is expensive, and crowd-sourced annotations may not be reliable. At the same time, recent work in machine learning has demonstrated that learning from soft-labels acquired from crowd annotations can be effective, especially when there is distribution shift in the test set. However, the best method for acquiring these soft labels is inconsistent across tasks. This paper proposes new methods for acquiring soft-labels from crowd-annotations by aggregating the distributions produced by existing methods. In particular, we propose to find a distribution over classes by learning from multiple-views of crowd annotations via temperature scaling and finding the Jensen-Shannon centroid of their distributions. We demonstrate that using these aggregation methods leads to best or near-best performance across four NLP tasks on out-of-domain test sets, mitigating fluctuations in performance when using the constituent methods on their own. Additionally, these methods result in best or near-best uncertainty estimation across tasks. We argue that aggregating different views of crowd-annotations as soft-labels is an effective way to ensure performance which is as good or better than the best individual view, which is useful given the inconsistency in performance of the individual methods.

The analysis of digitized historical manuscripts is typically addressed by paleographic experts. Writer identification refers to the classification of known writers while writer retrieval seeks to find the writer by means of image similarity in a dataset of images. While automatic writer identification/retrieval methods already provide promising results for many historical document types, papyri data is very challenging due to the fiber structures and severe artifacts. Thus, an important step for an improved writer identification is the preprocessing and feature sampling process. We investigate several methods and show that a good binarization is key to an improved writer identification in papyri writings. We focus mainly on writer retrieval using unsupervised feature methods based on traditional or self-supervised-based methods. It is, however, also comparable to the state of the art supervised deep learning-based method in the case of writer classification/re-identification.

This paper develops a clustering method that takes advantage of the sturdiness of model-based clustering, while attempting to mitigate some of its pitfalls. First, we note that standard model-based clustering likely leads to the same number of clusters per margin, which seems a rather artificial assumption for a variety of datasets. We tackle this issue by specifying a finite mixture model per margin that allows each margin to have a different number of clusters, and then cluster the multivariate data using a strategy game-inspired algorithm to which we call Reign-and-Conquer. Second, since the proposed clustering approach only specifies a model for the margins -- but leaves the joint unspecified -- it has the advantage of being partially parallelizable; hence, the proposed approach is computationally appealing as well as more tractable for moderate to high dimensions than a `full' (joint) model-based clustering approach. A battery of numerical experiments on artificial data indicate an overall good performance of the proposed methods in a variety of scenarios, and real datasets are used to showcase their application in practice.

We can protect user data privacy via many approaches, such as statistical transformation or generative models. However, each of them has critical drawbacks. On the one hand, creating a transformed data set using conventional techniques is highly time-consuming. On the other hand, in addition to long training phases, recent deep learning-based solutions require significant computational resources. In this paper, we propose PrivateSMOTE, a technique designed for competitive effectiveness in protecting cases at maximum risk of re-identification while requiring much less time and computational resources. It works by synthetic data generation via interpolation to obfuscate high-risk cases while minimizing data utility loss of the original data. Compared to multiple conventional and state-of-the-art privacy-preservation methods on 20 data sets, PrivateSMOTE demonstrates competitive results in re-identification risk. Also, it presents similar or higher predictive performance than the baselines, including generative adversarial networks and variational autoencoders, reducing their energy consumption and time requirements by a minimum factor of 9 and 12, respectively.

对AI的道德影响和值得信赖系统的设计的研究需要分析使用AI系统的方案，这与“用例”的软件工程概念和“预期目的”法律术语有关。但是，没有用于涵盖使用，范围，功能要求和AI系统风险的用例文档的标准方法。在这项工作中，我们为AI用例提出了一种新颖的文档方法，特别关注情感计算领域。我们的方法基于对研究文献中记录的用例信息需求的评估以及最近提议的AI欧洲监管框架。通过此评估，我们采用并调整了统一的建模语言（UML），在过去的二十年中，这主要由软件工程师使用。然后，每个用例都由UML图和一个结构化表表示，我们提供了一组示例，说明了其在几种情感计算方案中的应用。

目前，自然语言理解（NLU）中最根本的两个挑战是：（a）如何以“正确”的原因确定基于深度学习的模型是否在NLU基准上得分很高；（b）了解这些原因甚至是什么。我们研究了关于两个语言“技能”的阅读理解模型的行为：核心分辨率和比较。我们为从系统中预期的推理步骤提出了一个定义，该系统将“缓慢阅读”，并将其与各种大小的贝特家族的五个模型的行为进行比较，这是通过显着分数和反事实解释观察到的。我们发现，对于比较（而不是核心），基于较大编码器的系统更有可能依靠“正确”的信息，但即使他们在概括方面也很难，表明他们仍然学习特定的词汇模式，而不是比较的一般原则。