Search

In this paper, we investigate what types of stereotypical information are captured by pretrained language models. We present the first dataset comprising stereotypical attributes of a range of social groups and propose a method to elicit stereotypes encoded by pretrained language models in an unsupervised fashion. Moreover, we link the emergent stereotypes to their manifestation as basic emotions as a means to study their emotional effects in a more generalized manner. To demonstrate how our methods can be used to analyze emotion and stereotype shifts due to linguistic experience, we use fine-tuning on news sources as a case study. Our experiments expose how attitudes towards different social groups vary across models and how quickly emotions and stereotypes can shift at the fine-tuning stage.

A critical challenge faced by supervised word sense disambiguation (WSD) is the lack of large annotated datasets with sufficient coverage of words in their diversity of senses. This inspired recent research on few-shot WSD using meta-learning. While such work has successfully applied meta-learning to learn new word senses from very few examples, its performance still lags behind its fully supervised counterpart. Aiming to further close this gap, we propose a model of semantic memory for WSD in a meta-learning setting. Semantic memory encapsulates prior experiences seen throughout the lifetime of the model, which aids better generalization in limited data settings. Our model is based on hierarchical variational inference and incorporates an adaptive memory update rule via a hypernetwork. We show our model advances the state of the art in few-shot WSD, supports effective learning in extremely data scarce (e.g. one-shot) scenarios and produces meaning prototypes that capture similar senses of distinct words., 15 pages, 5 figures

Computational modelling of political discourse tasks has become an increasingly important area of research in natural language processing. Populist rhetoric has risen across the political sphere in recent years; however, computational approaches to it have been scarce due to its complex nature. In this paper, we present the new $\textit{Us vs. Them}$ dataset, consisting of 6861 Reddit comments annotated for populist attitudes and the first large-scale computational models of this phenomenon. We investigate the relationship between populist mindsets and social groups, as well as a range of emotions typically associated with these. We set a baseline for two tasks related to populist attitudes and present a set of multi-task learning models that leverage and demonstrate the importance of emotion and group identification as auxiliary tasks., Comment: Camera-ready version in EACL 2021

Philosophical reflection is traditionally presented in academic forms (as article, monograph, essay), but there is a rich tradition of philosophical reflection in literary genres. Drama can be understood as a literary genre, althoug there are the opposing views. In the twentieth century, philosophical reflection in drama became widespread.This phenomenon enriched the culture, changed the attitude to drama and expanded the understanding of how philosophical discourse can be carried out. The new Russian drama reflects the spiritual state peculiar to the era, and therefore it is interesting how philosophical reflection is carried out in the plays of modern authors. On the example of G. Zhuravleva's play, the typology of relations between Philosophy and Literature, proposed by D. K. Manokhin, is considered. In this typology, there are three models of relations between Philosophy and Literature: 1) the author appears in different texts as a philosopher and as a writer, 2) philosopher reflects on a literary text, 3) the writer imitates a philosophical text to achieve an artistic effect. The ideas of the author of the play are correlated with the philosophical heritage of the past (for example, with the ideas of Augustine "on the City of God", or K. Marx's concept of the basis and superstructure, or the concept of three types of "active life" by H. Arendt). It is concluded that the classification of variants of relations between philosophy and literature is problematic. The question is raised about the possibility of such a model of philosophical reflection, in which the author performs philosophical reflection in a dramatic, narrative way, while, in fact, performing work similar to that done by a philosopher in the traditional form of an academic text., Философское размышление традиционно представлено в академических формах (статья, монография, эссе), но существует богатая традиция философского размышления в литературных жанрах. В ХХ веке широкое распространение получило философское размышление в драме (которую можно понимать как литературный жанр, признавая наличие противоположных взглядов), что обогатило культуру, изменило отношение к драматургии и расширило понимание того, как может осуществляться философский дискурс. В новой российской драме нашло отражение духовное состояние, свойственное эпохе, в связи с чем интересно, как осуществляется философское размышление в пьесах современных авторов. На примере пьесы Г. Журавлевой рассмотрена типология отношений философии и литературы, предложенная Д.К. Манохиным. В этой типологии существует три модели отношений философии и литературы: 1) автор выступает в разных тексах как философ и как литератор, 2) автор - философ размышляет о литературном тексте, 3) писатель имитирует философский текст для достижения художественного эффекта. Идеи автора пьесы соотнесены с философским наследием прошлого (например, с идеями Августина «о Граде Божьем», концепцией К.Маркса о базисе и надстройке, концепцией трех видов «деятельной жизни» Х.Арендт). Сделан вывод о проблематичности классификации вариантов отношений между философией и литературой. Поставлен вопрос о возможности такой модели философского размышления, в которой автор совершает философское размышление драматургически, нарративным способом, при этом, фактически, совершая работу, аналогичную той, что делает философ в традиционной форме академического текста.

В наши дни актуальной становится проблема понимания золотого правила нравственности. В статье обсуждаются результаты исследования того, как молодые люди интерпретируют “золотое правило” нравственности. Выявлена потребность в объяснении базовых этических концепций, в целом положительное отношение к “золотому правилу”, склонность следовать собственным интересам в ситуации нравственного выбора. Today the problem related to the understanding of the “Golden Rule” of morality is increasing. The article considers the interpretation of the “Golden Rule” principle by young people. The results of the study show that young people lack an understanding of basic ethical concepts, however in general aware of the “Golden Rule”, have a positive view of this principle, but a tendency to follow their own interests in a situation of moral choice.

The article presents the results of an anthropometric study of the body of students of a special medical group of the Ural Federal University. Assessment of physical development was supplemented by evaluation indices derived by comparing different anthropometric attributes. Shown are the changes in the physical condition of students through the implementation of a program for water aerobics with elements of sports swimming. В статье представлены результаты антропометрического ис-следования организма студентов специальной медицинской группы Уральского федерального университета. Оценка физического развития дополнялась оценочными индексами, выводимыми путем сопоставления разных антропометрических признаков. Показаны изменения физического состояния студентов с помощью реализации программы по аквааэробике с элементами спортивного плавания.

The success of deep learning methods hinges on the availability of large training datasets annotated for the task of interest. In contrast to human intelligence, these methods lack versatility and struggle to learn and adapt quickly to new tasks, where labeled data is scarce. Meta-learning aims to solve this problem by training a model on a large number of few-shot tasks, with an objective to learn new tasks quickly from a small number of examples. In this paper, we propose a meta-learning framework for few-shot word sense disambiguation (WSD), where the goal is to learn to disambiguate unseen words from only a few labeled instances. Meta-learning approaches have so far been typically tested in an $N$-way, $K$-shot classification setting where each task has $N$ classes with $K$ examples per class. Owing to its nature, WSD deviates from this controlled setup and requires the models to handle a large number of highly unbalanced classes. We extend several popular meta-learning approaches to this scenario, and analyze their strengths and weaknesses in this new challenging setting., Camera-ready: Findings of EMNLP

The rise of online communication platforms has been accompanied by some undesirable effects, such as the proliferation of aggressive and abusive behaviour online. Aiming to tackle this problem, the natural language processing (NLP) community has experimented with a range of techniques for abuse detection. While achieving substantial success, these methods have so far only focused on modelling the linguistic properties of the comments and the online communities of users, disregarding the emotional state of the users and how this might affect their language. The latter is, however, inextricably linked to abusive behaviour. In this paper, we present the first joint model of emotion and abusive language detection, experimenting in a multi-task learning framework that allows one task to inform the other. Our results demonstrate that incorporating affective features leads to significant improvements in abuse detection performance across datasets.

We apply contextualised word embeddings to lexical semantic change detection in the SemEval-2020 Shared Task 1. This paper focuses on Subtask 2, ranking words by the degree of their semantic drift over time. We analyse the performance of two contextualising architectures (BERT and ELMo) and three change detection algorithms. We find that the most effective algorithms rely on the cosine similarity between averaged token embeddings and the pairwise distances between token embeddings. They outperform strong baselines by a large margin (in the post-evaluation phase, we have the best Subtask 2 submission for SemEval-2020 Task 1), but interestingly, the choice of a particular algorithm depends on the distribution of gold scores in the test set.

There has been an increased interest in modelling political discourse within the natural language processing (NLP) community, in tasks such as political bias and misinformation detection, among others. Metaphor-rich and emotion-eliciting communication strategies are ubiquitous in political rhetoric, according to social science research. Yet, none of the existing computational models of political discourse has incorporated these phenomena. In this paper, we present the first joint models of metaphor, emotion and political rhetoric, and demonstrate that they advance performance in three tasks: predicting political perspective of news articles, party affiliation of politicians and framing of policy issues.

In this study we investigate how lexical-semantic relations associated with the literal meaning (and abstract meaning) are being accessed across the brain during familiar metaphor comprehension. We utilize a data-driven whole-brain searchlight similarity-decoding analysis. We contrast decoding metaphoric phrases ("she's grasping the idea") using distributional semantic models of the verb in the phrase (Verb model) versus that of the more abstract verb-sense (Paraphrase Verb model) obtained from literal paraphrases of the metaphoric phrases ("she's understanding the idea"). We showed successful decoding with the Verb model across frontal, temporal and parietal lobes mainly within areas of the language and default-mode networks. In contrast, decoding with the Paraphrase Verb model was restricted to frontal-temporal lobes within areas of the language-network which overlapped to some extent with significant decoding with the Verb model. Overall, the results suggest that lexical-semantic relations closely associated with the abstract meaning in metaphor processing are largely localized to language and amodal (multimodal) semantic memory systems of the brain, while those more associated with the literal meaning are processed across a more distributed semantic network including areas implicated in mental imagery and social-cognition.

We present the Narrative Brain Dataset, an fMRI dataset that was collected during spoken presentation of short excerpts of three stories in Dutch. Together with the brain imaging data, the dataset contains the written versions of the stimulation texts. The texts are accompanied with stochastic (perplexity and entropy) and semantic computational linguistic measures. The richness and unconstrained nature of the data allows the study of language processing in the brain in a more naturalistic setting than is common for fMRI studies. We hope that by making NBD available we serve the double purpose of providing useful neural data to researchers interested in natural language processing in the brain and to further stimulate data sharing in the field of neuroscience of language.

Most current approaches to metaphor identification use restricted linguistic contexts, e.g. by considering only a verb's arguments or the sentence containing a phrase. Inspired by pragmatic accounts of metaphor, we argue that broader discourse features are crucial for better metaphor identification. We train simple gradient boosting classifiers on representations of an utterance and its surrounding discourse learned with a variety of document embedding methods, obtaining near state-of-the-art results on the 2018 VU Amsterdam metaphor identification task without the complex metaphor-specific features or deep neural architectures employed by other systems. A qualitative analysis further confirms the need for broader context in metaphor processing., Comment: NAACL 2019; 6 pages; code available at https://github.com/jayelm/broader-metaphor; v2 updates affiliations and acknowledgments

Когда в мире появляется что-то новое, люди возлагают на это очень большие надежды. Так и получилось с появлением относительно молодой науки — генетики. Но, как и любая наука, генетика несёт в себе трудности, не только в плане исследований, но и в плане восприятия её обществом. Совместимы ли этика и генетика? When something new appears in the world, people place great hopes on it. So it happened with the advent of young science — genetics. But like any science, genetics carries with it difficulties, not only in terms of research, but also in terms of its perception by society. Are ethics and genetics compatible?

Recent work shows that distributional semantic models can be used to decode patterns of brain activity associated with individual words and sentence meanings. However, it is yet unclear to what extent such models can be used to study and decode fMRI patterns associated with specific aspects of semantic composition such as the negation function. In this paper, we apply lexical and compositional semantic models to decode fMRI patterns associated with negated and affirmative sentences containing hand-action verbs. Our results show reduced decoding (correlation) of sentences where the verb is in the negated context, as compared to the affirmative one, within brain regions implicated in action-semantic processing. This supports behavioral and brain imaging studies, suggesting that negation involves reduced access to aspects of the affirmative mental representation. The results pave the way for testing alternate semantic models of negation against human semantic processing in the brain.

Abuse on the Internet represents a significant societal problem of our time. Previous research on automated abusive language detection in Twitter has shown that community-based profiling of users is a promising technique for this task. However, existing approaches only capture shallow properties of online communities by modeling follower–following relationships. In contrast, working with graph convolutional networks (GCNs), we present the first approach that captures not only the structure of online communities but also the linguistic behavior of the users within them. We show that such a heterogeneous graph-structured modeling of communities significantly advances the current state of the art in abusive language detection.

As the community working on computational approaches to figurative language is growing and as methods and data become increasingly diverse, it is important to create widely shared empirical knowledge of the level of system performance in a range of contexts, thus facilitating progress in this area. One way of creating such shared knowledge is through benchmarking multiple systems on a common dataset. We report on the shared task on metaphor identification on the VU Amsterdam Metaphor Corpus conducted at the NAACL 2018 Workshop on Figurative Language Processing.

The rapid growth of social media in recent years has fed into some highly undesirable phenomena such as proliferation of hateful and offensive language on the Internet. Previous research suggests that such abusive content tends to come from users who share a set of common stereotypes and form communities around them. The current state-of-the-art approaches to abuse detection are oblivious to user and community information and rely entirely on textual (i.e., lexical and semantic) cues. In this paper, we propose a novel approach to this problem that incorporates community-based profiling features of Twitter users. Experimenting with a dataset of 16k tweets, we show that our methods significantly outperform the current state of the art in abuse detection. Further, we conduct a qualitative analysis of model characteristics. We release our code, pre-trained models and all the resources used in the public domain.

The advent of social media in recent years has fed into some highly undesirable phenomena such as proliferation of offensive language, hate speech, sexist remarks, etc. on the Internet. In light of this, there have been several efforts to automate the detection and moderation of such abusive content. However, deliberate obfuscation of words by users to evade detection poses a serious challenge to the effectiveness of these efforts. The current state of the art approaches to abusive language detection, based on recurrent neural networks, do not explicitly address this problem and resort to a generic OOV (out of vocabulary) embedding for unseen words. However, in using a single embedding for all unseen words we lose the ability to distinguish between obfuscated and non-obfuscated or rare words. In this paper, we address this problem by designing a model that can compose embeddings for unseen words. We experimentally demonstrate that our approach significantly advances the current state of the art in abuse detection on datasets from two different domains, namely Twitter and Wikipedia talk page., Comment: In Proceedings of the EMNLP Workshop on Abusive Language Online 2018

One of the key problems in computational metaphor modelling is finding the optimal level of abstraction of semantic representations, such that these are able to capture and generalise metaphorical mechanisms. In this paper we present the first metaphor identification method that uses representations constructed from property norms. Such norms have been previously shown to provide a cognitively plausible representation of concepts in terms of semantic properties. Our results demonstrate that such property-based semantic representations provide a suitable model of cross-domain knowledge projection in metaphors, outperforming standard distributional models on a metaphor identification task.