Stories are rich in the emotions they exhibit in their narratives and evoke in the readers. The emotional journeys of the various characters within a story are central to their appeal. Computational analysis of the emotions of novels, however, has rarely examined the variation in the emotional trajectories of the different characters within them, instead considering the entire novel to represent a single story arc. In this work, we use character dialogue to distinguish between the emotion arcs of the narration and the various characters. We analyze the emotion arcs of the various characters in a dataset of English literary novels using the framework of Utterance Emotion Dynamics. Our findings show that the narration and the dialogue largely express disparate emotions through the course of a novel, and that the commonalities or differences in the emotional arcs of stories are more accurately captured by those associated with individual characters.

Current models for quotation attribution in literary novels assume varying levels of available information in their training and test data, which poses a challenge for in-the-wild inference. Here, we approach quotation attribution as a set of four interconnected sub-tasks: character identification, coreference resolution, quotation identification, and speaker attribution. We benchmark state-of-the-art models on each of these sub-tasks independently, using a large dataset of annotated coreferences and quotations in literary novels (the Project Dialogism Novel Corpus). We also train and evaluate models for the speaker attribution task in particular, showing that a simple sequential prediction model achieves accuracy scores on par with state-of-the-art models.

We present the Project Dialogism Novel Corpus, or PDNC, an annotated dataset of quotations for English literary texts. PDNC contains annotations for 35,978 quotations across 22 full-length novels, and is by an order of magnitude the largest corpus of its kind. Each quotation is annotated for the speaker, addressees, type of quotation, referring expression, and character mentions within the quotation text. The annotated attributes allow for a comprehensive evaluation of models of quotation attribution and coreference for literary texts.

Operationalizing morality is crucial for understanding multiple aspects of society that have moral values at their core – such as riots, mobilizing movements, public debates, etc. Moral Foundations Theory (MFT) has become one of the most adopted theories of morality partly due to its accompanying lexicon, the Moral Foundation Dictionary (MFD), which offers a base for computationally dealing with morality. In this work, we exploit the MFD in a novel direction by investigating how well moral values are captured by KGs. We explore three widely used KGs, and provide concept-level analogues for the MFD. Furthermore, we propose several Personalized PageRank variations in order to score all the concepts and entities in the KGs with respect to their relevance to the different moral values. Our promising results help to progress the operationalization of morality in both NLP and KG communities.

We present a text-based framework for investigating moral sentiment change of the public via longitudinal corpora. Our framework is based on the premise that language use can inform people’s moral perception toward right or wrong, and we build our methodology by exploring moral biases learned from diachronic word embeddings. We demonstrate how a parameter-free model supports inference of historical shifts in moral sentiment toward concepts such as slavery and democracy over centuries at three incremental levels: moral relevance, moral polarity, and fine-grained moral dimensions. We apply this methodology to visualizing moral time courses of individual concepts and analyzing the relations between psycholinguistic variables and rates of moral sentiment change at scale. Our work offers opportunities for applying natural language processing toward characterizing moral sentiment change in society.

Word embeddings are often criticized for capturing undesirable word associations such as gender stereotypes. However, methods for measuring and removing such biases remain poorly understood. We show that for any embedding model that implicitly does matrix factorization, debiasing vectors post hoc using subspace projection (Bolukbasi et al., 2016) is, under certain conditions, equivalent to training on an unbiased corpus. We also prove that WEAT, the most common association test for word embeddings, systematically overestimates bias. Given that the subspace projection method is provably effective, we use it to derive a new measure of association called the relational inner product association (RIPA). Experiments with RIPA reveal that, on average, skipgram with negative sampling (SGNS) does not make most words any more gendered than they are in the training corpus. However, for gender-stereotyped words, SGNS actually amplifies the gender association in the corpus.

A surprising property of word vectors is that word analogies can often be solved with vector arithmetic. However, it is unclear why arithmetic operators correspond to non-linear embedding models such as skip-gram with negative sampling (SGNS). We provide a formal explanation of this phenomenon without making the strong assumptions that past theories have made about the vector space and word distribution. Our theory has several implications. Past work has conjectured that linear substructures exist in vector spaces because relations can be represented as ratios; we prove that this holds for SGNS. We provide novel justification for the addition of SGNS word vectors by showing that it automatically down-weights the more frequent word, as weighting schemes do ad hoc. Lastly, we offer an information theoretic interpretation of Euclidean distance in vector spaces, justifying its use in capturing word dissimilarity.

According to the literary theory of Mikhail Bakhtin, a dialogic novel is one in which characters speak in their own distinct voices, rather than serving as mouthpieces for their authors. We use text classification to determine which authors best achieve dialogism, looking at a corpus of plays from the late nineteenth and early twentieth centuries. We find that the SAGE model of text generation, which highlights deviations from a background lexical distribution, is an effective method of weighting the words of characters’ utterances. Our results show that it is indeed possible to distinguish characters by their speech in the plays of canonical writers such as George Bernard Shaw, whereas characters are clustered more closely in the works of lesser-known playwrights.

We present two models for combining word and character embeddings for cause-of-death classification of verbal autopsy reports using the text of the narratives. We find that for smaller datasets (500 to 1000 records), adding character information to the model improves classification, making character-based CNNs a promising method for automated verbal autopsy coding.

We introduce a multi-task learning model for cause-of-death classification of verbal autopsy narratives that jointly learns to output interpretable key phrases. Adding these key phrases outperforms the baseline model and topic modeling features.

We created a corpus of utterances that attempt to save face from parliamentary debates and use it to automatically analyze the language of reputation defence. Our proposed model that incorporates information regarding threats to reputation can predict reputation defence language with high confidence. Further experiments and evaluations on different datasets show that the model is able to generalize to new utterances and can predict the language of reputation defence in a new dataset.

We present an automated approach to distinguish true, false, stretch, and dodge statements in questions and answers in the Canadian Parliament. We leverage the truthfulness annotations of a U.S. fact-checking corpus by training a neural net model and incorporating the prediction probabilities into our models. We find that in concert with other linguistic features, these probabilities can improve the multi-class classification results. We further show that dodge statements can be detected with an F1 measure as high as 82.57% in binary classification settings.

We present metrics for listwise temporal ordering of events in clinical notes, as well as a baseline listwise temporal ranking model that generates a timeline of events that can be used in downstream medical natural language processing tasks.

Research on computational argumentation faces the problem of how to automatically assess the quality of an argument or argumentation. While different quality dimensions have been approached in natural language processing, a common understanding of argumentation quality is still missing. This paper presents the first holistic work on computational argumentation quality in natural language. We comprehensively survey the diverse existing theories and approaches to assess logical, rhetorical, and dialectical quality dimensions, and we derive a systematic taxonomy from these. In addition, we provide a corpus with 320 arguments, annotated for all 15 dimensions in the taxonomy. Our results establish a common ground for research on computational argumentation quality assessment.

Current approaches to cross-lingual sentiment analysis try to leverage the wealth of labeled English data using bilingual lexicons, bilingual vector space embeddings, or machine translation systems. Here we show that it is possible to use a single linear transformation, with as few as 2000 word pairs, to capture fine-grained sentiment relationships between words in a cross-lingual setting. We apply these cross-lingual sentiment models to a diverse set of tasks to demonstrate their functionality in a non-English context. By effectively leveraging English sentiment knowledge without the need for accurate translation, we can analyze and extract features from other languages with scarce data at a very low cost, thus making sentiment and related analyses for many languages inexpensive.

Argumentation quality is viewed differently in argumentation theory and in practical assessment approaches. This paper studies to what extent the views match empirically. We find that most observations on quality phrased spontaneously are in fact adequately represented by theory. Even more, relative comparisons of arguments in practice correlate with absolute quality ratings based on theory. Our results clarify how the two views can learn from each other.

We propose a new task of automatically detecting reputation defence strategies in the field of computational argumentation. We cast the problem as relation classification, where given a pair of reputation threat and reputation defence, we determine the reputation defence strategy. We annotate a dataset of parliamentary questions and answers with reputation defence strategies. We then propose a model based on supervised learning to address the detection of these strategies, and report promising experimental results.

Previous approaches to generic frame classification analyze frames at the document level. Here, we propose a supervised based approach based on deep neural networks and distributional representations for classifying frames at the sentence level in news articles. We conduct our experiments on the publicly available Media Frames Corpus compiled from the U.S. Newspapers. Using (B)LSTMs and GRU networks to represent the meaning of frames, we demonstrate that our approach yields at least 14-point improvement over several baseline methods.

T. S. Eliot’s poem The Waste Land is a notoriously challenging example of modernist poetry, mixing the independent viewpoints of over ten distinct characters without any clear demarcation of which voice is speaking when. In this work, we apply unsupervised techniques in computational stylistics to distinguish the particular styles of these voices, offering a computer’s perspective on longstanding debates in literary analysis. Our work includes a model for stylistic segmentation that looks for points of maximum stylistic variation, a k-means clustering model for detecting non-contiguous speech from the same voice, and a stylistic profiling approach which makes use of lexical resources built from a much larger collection of literary texts. Evaluating using an expert interpretation, we show clear progress in distinguishing the voices of The Waste Land as compared to appropriate baselines, and we also offer quantitative evidence both for and against that particular interpretation.

The task of native language (L1) identification suffers from a relative paucity of useful training corpora, and standard within-corpus evaluation is often problematic due to topic bias. In this paper, we introduce a method for L1 identification in second language (L2) texts that relies only on much more plentiful L1 data, rather than the L2 texts that are traditionally used for training. In particular, we do word-by-word translation of large L1 blog corpora to create a mapping to L2 forms that are a possible result of language transfer, and then use that information for unsupervised classification. We show this method is effective in several different learner corpora, with bigram features being particularly useful.