_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d222140918 | A frequent pattern in customer care conversations is the agents responding with appropriate webpage URLs that address users' needs. We study the task of predicting the documents that customer care agents can use to facilitate users' needs. We also introduce a new public dataset 1 which supports the aforementioned problem. Using this dataset and two others, we investigate state-of-the-art deep learning (DL) and information retrieval (IR) models for the task. We also analyze the practicality of such systems in terms of inference time complexity. Our results show that an hybrid IR+DL approach provides the best of both worlds. * Work done when author was at IBM Research 1 The Twitter dataset is available at: https://github.com/IBM/twitter-customercare-document-prediction 2 The terms URL and document are used interchangeably. | Conversational Document Prediction to Assist Customer Care Agents |
d18941595 | In this paper the author presents methods for dynamic terminology integration in statistical machine translation systems using a source text pre-processing workflow. The workflow consists of exchangeable components for term identification, inflected form generation for terms, and term translation candidate ranking. Automatic evaluation for three language pairs shows a translation quality improvement from 0.9 to 3.41 BLEU points over the baseline. Manual evaluation for seven language pairs confirms the positive results; the proportion of correctly translated terms increases from 1.6% to 52.6% over the baseline. | Dynamic Terminology Integration Methods in Statistical Machine Translation |
d235097529 | The success of pretrained word embeddings has motivated their use in the biomedical domain, with contextualized embeddings yielding remarkable results in several biomedical NLP tasks. However, there is a lack of research on quantifying their behavior under severe "stress" scenarios. In this work, we systematically evaluate three language models with adversarial examples -automatically constructed tests that allow us to examine how robust the models are. We propose two types of stress scenarios focused on the biomedical named entity recognition (NER) task, one inspired by spelling errors and another based on the use of synonyms for medical terms. Our experiments with three benchmarks show that the performance of the original models decreases considerably, in addition to revealing their weaknesses and strengths. Finally, we show that adversarial training causes the models to improve their robustness and even to exceed the original performance in some cases. | Stress Test Evaluation of Biomedical Word Embeddings |
d13362841 | General treebank analyses are graph structured, but parsers are typically restricted to tree structures for efficiency and modeling reasons. We propose a new representation and algorithm for a class of graph structures that is flexible enough to cover almost all treebank structures, while still admitting efficient learning and inference. In particular, we consider directed, acyclic, one-endpoint-crossing graph structures, which cover most long-distance dislocation, shared argumentation, and similar tree-violating linguistic phenomena. We describe how to convert phrase structure parses, including traces, to our new representation in a reversible manner. Our dynamic program uniquely decomposes structures, is sound and complete, and covers 97.3% of the Penn English Treebank. We also implement a proofof-concept parser that recovers a range of null elements and trace types. | Parsing with Traces: An O(n 4 ) Algorithm and a Structural Representation |
d12412559 | Zoonotic viruses represent emerging or re-emerging pathogens that pose significant public health threats throughout the world. It is therefore crucial to advance current surveillance mechanisms for these viruses through outlets such as phylogeography. Despite the abundance of zoonotic viral sequence data in publicly available databases such as GenBank, phylogeographic analysis of these viruses is often limited by the lack of adequate geographic metadata. However, many GenBank records include references to articles with more detailed information and automated systems may help extract this information efficiently and effectively. In this paper, we describe our efforts to determine the proportion of GenBank records with "insufficient" geographic metadata for seven well-studied viruses. We also evaluate the performance of four different Named Entity Recognition (NER) systems for automatically extracting related entities using a manually created gold-standard. | Natural Language Processing Methods for Enhancing Geographic Metadata for Phylogeography of Zoonotic Viruses |
d214774853 | When does a sequence of events define an everyday scenario and how can this knowledge be induced from text? Prior works in inducing such scripts have relied on, in one form or another, measures of correlation between instances of events in a corpus. We argue from both a conceptual and practical sense that a purely correlation-based approach is insufficient, and instead propose an approach to script induction based on the causal effect between events, formally defined via interventions. Through both human and automatic evaluations, we show that the output of our method based on causal effects better matches the intuition of what a script represents. | Causal Inference of Script Knowledge |
d248779985 | Self-supervised models for speech processing form representational spaces without using any external labels. Increasingly, they appear to be a feasible way of at least partially eliminating costly manual annotations, a problem of particular concern for low-resource languages. But what kind of representational spaces do these models construct? Human perception specializes to the sounds of listeners' native languages. Does the same thing happen in self-supervised models? We examine the representational spaces of three kinds of stateof-the-art self-supervised models: wav2vec 2.0, HuBERT and contrastive predictive coding (CPC), and compare them with the perceptual spaces of French-speaking and Englishspeaking human listeners, both globally and taking account of the behavioural differences between the two language groups. We show that the CPC model shows a small native language effect, but that wav2vec 2.0 and Hu-BERT seem to develop a universal speech perception space which is not language specific. A comparison against the predictions of supervised phone recognisers suggests that all three self-supervised models capture relatively finegrained perceptual phenomena, while supervised models are better at capturing coarser, phone-level, effects of listeners' native language, on perception. . 2020. wav2vec 2.0: A framework for self-supervised learning of speech representations. arXiv preprint arXiv:2006.11477.MP Cooke and OE Scharenborg. 2008. The interspeech 2008 consonant challenge. | Do self-supervised speech models develop human-like perception biases? |
d254877577 | We present, Naamapadam, the largest publicly available Named Entity Recognition (NER) dataset for the 11 major Indian languages from two language families. The dataset contains more than 400k sentences annotated with a total of at least 100k entities from three standard entity categories (Person, Location, and, Organization) for 9 out of the 11 languages. The training dataset has been automatically created from the Samanantar parallel corpus by projecting automatically tagged entities from an English sentence to the corresponding Indian language translation. We also create manually annotated testsets for 9 languages. We demonstrate the utility of the obtained dataset on the Naamapadam-test dataset. We also release In-dicNER, a multilingual IndicBERT model finetuned on Naamapadam training set. IndicNER achieves an F1 score of more than 80 for 7 out of 9 test languages. The dataset and models are available under open-source licences at https: //ai4bharat.iitm.ac.in/naamapadam. * Equal contribution † Project leads; correspondence to rmurthyv@in.ibm.com, | Naamapadam: A Large-Scale Named Entity Annotated Data for Indic Languages |
d248392209 | In this work, we propose a flow-adapter architecture for unsupervised NMT. It leverages normalizing flows to explicitly model the distributions of sentence-level latent representations, which are subsequently used in conjunction with the attention mechanism for the translation task. The primary novelties of our model are: (a) capturing language-specific sentence representations separately for each language using normalizing flows and (b) using a simple transformation of these latent representations for translating from one language to another. This architecture allows for unsupervised training of each language independently. While there is prior work on latent variables for supervised MT, to the best of our knowledge, this is the first work that uses latent variables and normalizing flows for unsupervised MT. We obtain competitive results on several unsupervised MT benchmarks. | Flow-Adapter Architecture for Unsupervised Machine Translation |
d7142721 | In this paper, we describe an annotation scheme for the attribution of abstract objects (propositions, facts, and eventualities) associated with discourse relations and their arguments annotated in the Penn Discourse TreeBank. The scheme aims to capture both the source and degrees of factuality of the abstract objects through the annotation of text spans signalling the attribution, and of features recording the source, type, scopal polarity, and determinacy of attribution.RÉSUMÉ. Dans cet article, nous décrivons un schéma d'annotation pour l'encodage des objets abstraits (propositions, faits et possibilités) associés aux relations de discours et à leurs arguments tels qu'annotés dans le Penn Discourse TreeBank. Ce schéma a pour objet la capture de la source et du degré de factualité des objets abstraits. Les aspects clés de ce schéma comprennent l'annotation des intervalles textuels signalant l'attribution, ainsi que l'annotation des proprietés caractérisant la source, le type, la polarité de la portée, et le degré de détermination de l'attribution. | Edinburgh Research Explorer Attribution and its Annotation in the Penn Discourse TreeBank Attribution and its annotation in the Penn Discourse TreeBank |
d261342039 | Animal words carry a large number of cognitive mappings of human society, and different ethnic groups have similarities and differences in cognition of the same word. The study of cognitive differences in animal words through metaphor is a very popular trend in recent years, and reflecting the cognitive attributes of people's cognitive impressions of words is a simple entry point. In this paper, 54 animals in the "Cognitive Attribute Database of Traditional Chinese Cultural Terms" are selected and the cognitive attribute differences in English and Chinese languages are compared and analyzed with the help of Chinese and English cognitive attribute databases. This paper finds that there are obvious differences between the cognitive attributes of animal words in English and Chinese, and the differences are more manifested in subjective attributes, and the overall similarities and differences between the cognitive attributes of animal words in Chinese and English are found. | Quantitative studies of cognitive attributes of English and Chinese animal words |
d258480179 | Keeping track of how states of entities change as a text or dialog unfolds is a key prerequisite to discourse understanding. Yet, there have been few systematic investigations into the ability of large language models (LLMs) to track discourse entities. In this work, we present a task probing to what extent a language model can infer the final state of an entity given an English description of the initial state and a series of state-changing operations. We use this task to first investigate whether Flan-T5, GPT-3 and GPT-3.5 can track the state of entities, and find that only GPT-3.5 models, which have been pretrained on large amounts of code, exhibit this ability. We then investigate whether smaller models pretrained primarily on text can learn to track entities, through finetuning T5 on several training/evaluation splits. While performance degrades for more complex splits, we find that even when evaluated on a different set of entities from training or longer operation sequences, a finetuned model can perform nontrivial entity tracking. Taken together, these results suggest that language models can learn to track entities but pretraining on text corpora alone does not make this capacity surface. | Entity Tracking in Language Models |
d258557186 | Large language models (LLMs) have demonstrated an impressive ability to generate codes on competitive programming tasks. However, with limited sample numbers, LLMs still suffer from poor accuracy. Inspired by the process of human programming, we propose a generateand-edit approach named Self-Edit that utilizes execution results of the generated code from LLMs to improve the code quality on the competitive programming task. We execute the generated code on the example test case provided in the question and wrap execution results into a supplementary comment. Utilizing this comment as guidance, our fault-aware code editor is employed to correct errors in the generated code. We perform extensive evaluations across two competitive programming datasets with nine different LLMs. Compared to directly generating from LLMs, our approach can improve the average of pass@1 by 89% on APPS-dev, 31% on APPS-test, and 48% on HumanEval over nine popular code generation LLMs with parameter sizes ranging from 110M to 175B. Compared to other post-processing methods, our method demonstrates superior accuracy and efficiency. . 2022. Incoder:A generative model for code infilling and synthesis. CoRR, abs/2204.05999. . 2022. Coderl: Mastering code generation through pretrained models and deep reinforcement learning. In NeurIPS. . 2023a. Enabling programming thinking in large language models toward code generation. arXiv preprint arXiv:2305.06599. . 2022a. Codeeditor: Learning to edit source code with pre-trained models. arXiv preprint arXiv:2210.17040. | Self-Edit: Fault-Aware Code Editor for Code Generation |
d211146690 | Many sequence-to-sequence generation tasks, including machine translation and text-tospeech, can be posed as estimating the density of the output y given the input x: p(y|x). Given this interpretation, it is natural to evaluate sequence-to-sequence models using conditional log-likelihood on a test set. However, the goal of sequence-to-sequence generation (or structured prediction) is to find the best outputŷ given an input x, and each task has its own downstream metric R that scores a model output by comparing against a set of references y * : R(ŷ, y * |x). While we hope that a model that excels in density estimation also performs well on the downstream metric, the exact correlation has not been studied for sequence generation tasks. In this paper, by comparing several density estimators on five machine translation tasks, we find that the correlation between rankings of models based on log-likelihood and BLEU varies significantly depending on the range of the model families being compared. First, log-likelihood is highly correlated with BLEU when we consider models within the same family (e.g. autoregressive models, or latent variable models with the same parameterization of the prior). However, we observe no correlation between rankings of models across different families:(1) among non-autoregressive latent variable models, a flexible prior distribution is better at density estimation but gives worse generation quality than a simple prior, and (2) autoregressive models offer the best translation performance overall, while latent variable models with a normalizing flow prior give the highest held-out log-likelihood across all datasets. | On the Discrepancy between Density Estimation and Sequence Generation |
d249018130 | Children acquiring English make systematic errors on subject control sentences even after they have reached near-adult competence(Chomsky, 1969), possibly due to heuristics based on semantic roles(Maratsos, 1974). Given the advanced fluency of large generative language models, we ask whether model outputs are consistent with these heuristics, and to what degree different models are consistent with each other. We find that models can be categorized by behavior into three separate groups, with broad differences between the groups. The outputs of models in the largest group are consistent with positional heuristics that succeed on subject control but fail on object control. This result is surprising, given that object control is orders of magnitude more frequent in the text data used to train such models. We examine to what degree the models are sensitive to prompting with agent-patient information, finding that raising the salience of agent and patient relations results in significant changes in the outputs of most models. Based on this observation, we leverage an existing dataset of semantic proto-role annotations(White et al., 2020)to explore the connections between control and labeling event participants with properties typically associated with agents and patients. | The Curious Case of Control |
d7180947 | We present an approach to disambiguating verb senses which differ w.r.t. the inferences they allow. It combines standard ontological tools and formalisms with a formal semantic analysis and is hence more formalised and more detailed than existing lexical semantic resources like WordNet and FrameNet [Fellbaum, 1998, Baker et al., 1998]. The resource presented here implements formal semantic descriptions of verbs in the Web Ontology Language (OWL) and exploits its reasoning potential based on Description Logics (DL) for the disambiguation of verbs in context, since before the correct sense of a verb can be reliably determined, its syntactic arguments have to be disambiguated first. We present details on this process, which is based on a mapping from the French EuroWordNet [Vossen, 1998] to SUMO [Niles and Pease, 2003]. Moreover, we focus on the selectional restrictions of verbs w.r.t. the ontological type of their arguments, as well as their representation as necessary and sufficient conditions in the TBox. After a DL reasoner has identified the verb sense on the basis of these conditions, we make use of the more expressive Semantic Web Rule Language to calculate the inferences that are permitted on the selected interpretation. | Disambiguation of Polysemous Verbs for Rule-based Inferencing |
d237452272 | Human dialogue contains evolving concepts, and speakers naturally associate multiple concepts to compose a response. However, current dialogue models with the seq2seq framework lack the ability to effectively manage concept transitions and can hardly introduce multiple concepts to responses in a sequential decoding manner. To facilitate a controllable and coherent dialogue, in this work, we devise a conceptguided non-autoregressive model (CG-nAR) for open-domain dialogue generation. The proposed model comprises a multi-concept planning module that learns to identify multiple associated concepts from a concept graph and a customized Insertion Transformer that performs concept-guided non-autoregressive generation to complete a response. The experimental results on two public datasets show that CG-nAR can produce diverse and coherent responses, outperforming state-of-the-art baselines in both automatic and human evaluations with substantially faster inference speed. | Thinking Clearly, Talking Fast: Concept-Guided Non-Autoregressive Generation for Open-Domain Dialogue Systems |
d204788606 | One of the basic tasks of computational language documentation (CLD) is to identify word boundaries in an unsegmented phonemic stream. While several unsupervised monolingual word segmentation algorithms exist in the literature, they are challenged in real-world CLD settings by the small amount of available data. A possible remedy is to take advantage of glosses or translation in a foreign, wellresourced, language, which often exist for such data. In this paper, we explore and compare ways to exploit neural machine translation models to perform unsupervised boundary detection with bilingual information, notably introducing a new loss function for jointly learning alignment and segmentation. We experiment with an actual under-resourced language, Mboshi, and show that these techniques can effectively control the output segmentation length.4In this case, α ij = exp(e ij /T ) J k=1 exp(e ik /T ) . 5 A temperature below 1 would conversely sharpen the alignment distribution. We did not observe significant changes in segmentation performance varying the temperature parameter. | Controlling Utterance Length in NMT-based Word Segmentation with Attention |
d174800443 | The performance of Part-of-Speech tagging varies significantly across the treebanks of the Universal Dependencies project. This work points out that these variations may result from divergences between the annotation of train and test sets. We show how the annotation variation principle, introduced by Dickinson and Meurers(2003)to automatically detect errors in gold standard, can be used to identify inconsistencies between annotations ; we also evaluate their impact on prediction performance. | How Bad are PoS Taggers in Cross-Corpora Settings? Evaluating Annotation Divergence in the UD Project |
d258187507 | While quality estimation (QE) can play an important role in the translation process, its effectiveness relies on the availability and quality of training data. For QE in particular, high-quality labeled data is often lacking due to the high cost and effort associated with labeling such data. Aside from the data scarcity challenge, QE models should also be generalizable; i.e., they should be able to handle data from different domains, both generic and specific. To alleviate these two main issues -data scarcity and domain mismatch -this paper combines domain adaptation and data augmentation in a robust QE system. Our method first trains a generic QE model and then fine-tunes it on a specific domain while retaining generic knowledge. Our results show a significant improvement for all the language pairs investigated, better cross-lingual inference, and a superior performance in zero-shot learning scenarios as compared to state-of-the-art baselines. | Tailoring Domain Adaptation for Machine Translation Quality Estimation |
d235417177 | Maintaining consistent personas is essential for dialogue agents. Although tremendous advancements have been brought, the limitedscale of annotated persona-dense data are still barriers towards training robust and consistent persona-based dialogue models. In this work, we show how the challenges can be addressed by disentangling persona-based dialogue generation into two sub-tasks with a novel BERTover-BERT (BoB) model. Specifically, the model consists of a BERT-based encoder and two BERT-based decoders, where one decoder is for response generation, and another is for consistency understanding. In particular, to learn the ability of consistency understanding from large-scale non-dialogue inference data, we train the second decoder in an unlikelihood manner. Under different limited data settings, both automatic and human evaluations demonstrate that the proposed model outperforms strong baselines in response quality and persona consistency. | BoB: BERT Over BERT for Training Persona-based Dialogue Models from Limited Personalized Data |
d1977529 | Information extraction (IE) from text has largely focused on relations between individual entities, such as who has won which award. However, some facts are never fully mentioned, and no IE method has perfect recall. Thus, it is beneficial to also tap contents about the cardinalities of these relations, for example, how many awards someone has won. We introduce this novel problem of extracting cardinalities and discuss specific challenges that set it apart from standard IE. We present a distant supervision method using conditional random fields. A preliminary evaluation results in precision between 3% and 55%, depending on the difficulty of relations. | Cardinal Virtues: Extracting Relation Cardinalities from Text |
d237593027 | Bias is pervasive in NLP models, motivating the development of automatic debiasing techniques. Evaluation of NLP debiasing methods has largely been limited to binary attributes in isolation, e.g., debiasing with respect to binary gender or race, however many corpora involve multiple such attributes, possibly with higher cardinality. In this paper we argue that a truly fair model must consider 'gerrymandering' groups which comprise not only single attributes, but also intersectional groups. We evaluate a form of bias-constrained model which is new to NLP, as well an extension of the iterative nullspace projection technique which can handle multiple protected attributes. | Evaluating Debiasing Techniques for Intersectional Biases |
d43943649 | Sentiment analysis in low-resource languages suffers from a lack of annotated corpora to estimate high-performing models. Machine translation and bilingual word embeddings provide some relief through cross-lingual sentiment approaches. However, they either require large amounts of parallel data or do not sufficiently capture sentiment information. We introduce Bilingual Sentiment Embeddings (BLSE), which jointly represent sentiment information in a source and target language. This model only requires a small bilingual lexicon, a source-language corpus annotated for sentiment, and monolingual word embeddings for each language. We perform experiments on three language combinations (Spanish, Catalan, Basque) for sentencelevel cross-lingual sentiment classification and find that our model significantly outperforms state-of-the-art methods on four out of six experimental setups, as well as capturing complementary information to machine translation. Our analysis of the resulting embedding space provides evidence that it represents sentiment information in the resource-poor target language without any annotated data in that language. | Bilingual Sentiment Embeddings: Joint Projection of Sentiment Across Languages |
d219310118 | Over the last decade or so, it has become increasingly clear to many cognitive scientists that research into human language (and cognition in general, for that matter) has largely neglected how language and thought are embedded in the body and the world. As argued by, for instance, Clark (1997), cognition is fundamentally embodied, that is, it can only be studied in relation to human action, perception, thought, and experience. As Feldman puts it: "Human language and thought are crucially shaped by the properties of our bodies and the structure of our physical and social environment. Language and thought are not best studied as formal mathematics and logic, but as adaptations that enable creatures like us to thrive in a wide range of situations" (p. 7). Although it may seem paradoxical to try formalizing this view in a computational theory of language comprehension, this is exactly what From Molecule to Metaphor does. Starting from the assumption that human thought is neural computation, Feldman develops a computational theory that takes the embodied nature of language into account: the neural theory of language.The book comprises 27 short chapters, distributed over nine parts. Part I presents the basic ideas behind embodied language and cognition and explains how the embodiment of language is apparent in the brain: The neural circuits involved in a particular experience or action are, for a large part, the same circuits involved in processing language about this experience or action.Part II discusses neural computation, starting from the molecules that take part in information processing by neurons. This detailed exposition is followed by a description of neuronal networks in the human body, in particular in the brain.The description of the neural theory of language begins in Part III, where it is explained how localist neural networks, often used as psycholinguistic models, can represent the meaning of concepts. This is done by introducing triangle nodes into the network. Each triangle node connects the nodes representing a concept, a role, and a filler-for example, "pea," "has-color," and "green." Such networks are trained by a process called recruitment learning, which is described only very informally. This is certainly an interesting idea for combining propositional and connectionist models, but it does leave the reader with a number of questions. For instance, how is the concept distinguished from the filler when they can be interchanged, as in "cats, feed-on, mice" versus "mice, feed-on, cats." And on a more philosophical note: Where does this leave embodiment? The idea that there exists a node representing the concept "pea," neurally distinct from its properties and from experiences with peas, seems to introduce abstract and arbitrary symbols. These are quite alien to embodied theories of cognition, which generally assume modal and analogical perceptual symbols (Barsalou 1999) or even no symbols at all(Brooks 1991). | From Molecule to Metaphor: A Neural Theory of Language |
d249455064 | We introduce ReFinED, an efficient end-to-end entity linking model which uses fine-grained entity types and entity descriptions to perform linking. The model performs mention detection, fine-grained entity typing, and entity disambiguation for all mentions within a document in a single forward pass, making it more than 60 times faster than competitive existing approaches. ReFinED also surpasses state-ofthe-art performance on standard entity linking datasets by an average of 3.7 F1. The model is capable of generalising to large-scale knowledge bases such as Wikidata (which has 15 times more entities than Wikipedia) and of zeroshot entity linking. The combination of speed, accuracy and scale makes ReFinED an effective and cost-efficient system for extracting entities from web-scale datasets, for which the model has been successfully deployed. Our code and pre-trained models are available at https://github.com/alexa/ReFinED. | ReFinED: An Efficient Zero-shot-capable Approach to End-to-End Entity Linking |
d258480291 | As pointed out by several scholars, current research on hate speech (HS) recognition is characterized by unsystematic data creation strategies and diverging annotation schemata. Subsequently, supervised-learning models tend to generalize poorly to datasets they were not trained on, and the performance of the models trained on datasets labeled using different HS taxonomies cannot be compared. To ease this problem, we propose to apply extremely weak supervision that only relies on the class name rather than on class samples from the annotated data. We demonstrate the effectiveness of a state-of-the-art weakly-supervised text classification model in various in-dataset and crossdataset settings. Furthermore, we conduct an in-depth quantitative and qualitative analysis of the source of poor generalizability of HS classification models.Content Warning: This document discusses examples of harmful content (hate, abuse, and negative stereotypes). The authors do not support the use of harmful language. | Towards Weakly-Supervised Hate Speech Classification Across Datasets |
d259108873 | Misogyny and sexism are growing problems in social media. Advances have been made in online sexism detection but the systems are often uninterpretable. SemEval-2023 Task 10 on Explainable Detection of Online Sexism aims at increasing explainability of the sexism detection, and our team participated in all the proposed subtasks. Our system is based on further domain-adaptive pre-training(Gururangan et al., 2020). Building on the Transformerbased models with the domain adaptation, we compare fine-tuning with multi-task learning and show that each subtask requires a different system configuration. In our experiments, multi-task learning performs on par with standard fine-tuning for sexism detection and noticeably better for coarse-grained sexism classification, while fine-tuning is preferable for fine-grained classification 1 . | LCT-1 at SemEval-2023 Task 10: Pre-training and Multi-task Learning for Sexism Detection and Classification |
d5794284 | The NTT Statistical Machine Translation System consists of two primary components: a statistical machine translation decoder and a reranker. The decoder generates kbest translation canditates using a hierarchical phrase-based translation based on synchronous context-free grammar. The decoder employs a linear feature combination among several real-valued scores on translation and language models. The reranker reorders the k-best translation candidates using Ranking SVMs with a large number of sparse features. This paper describes the two components and presents the results for the evaluation campaign of IWSLT 2008. | NTT Statistical Machine Translation System for IWSLT 2008 |
d216651425 | In this paper, we give a treatment to the problem of bilingual part-of-speech induction with parallel data. We demonstrate that naïve optimization of log-likelihood with joint MRFs suffers from a severe problem of local maxima, and suggest an alternative -using contrastive estimation for estimation of the parameters. Our experiments show that estimating the parameters this way, using overlapping features with joint MRFs performs better than previous work on the 1984 dataset. | Unsupervised Bilingual POS Tagging with Markov Random Fields |
d9572199 | Mining bilingual data (including bilingual sentences and terms 1 ) from the Web can benefit many NLP applications, such as machine translation and cross language information retrieval. In this paper, based on the observation that bilingual data in many web pages appear collectively following similar patterns, an adaptive pattern-based bilingual data mining method is proposed. Specifically, given a web page, the method contains four steps: 1) preprocessing: parse the web page into a DOM tree and segment the inner text of each node into snippets; 2) seed mining: identify potential translation pairs (seeds) using a word based alignment model which takes both translation and transliteration into consideration; 3) pattern learning: learn generalized patterns with the identified seeds; 4) pattern based mining: extract all bilingual data in the page using the learned patterns. Our experiments on Chinese web pages produced more than 7.5 million pairs of bilingual sentences and more than 5 million pairs of bilingual terms, both with over 80% accuracy. | Mining Bilingual Data from the Web with Adaptively Learnt Patterns |
d5880140 | This article introduces a novel transition system for discontinuous lexicalized constituent parsing called SR-GAP. | Incremental Discontinuous Phrase Structure Parsing with the GAP Transition |
d49668568 | Multimodal machine learning is a core research area spanning the language, visual and acoustic modalities. The central challenge in multimodal learning involves learning representations that can process and relate information from multiple modalities. In this paper, we propose two methods for unsupervised learning of joint multimodal representations using sequence to sequence (Seq2Seq) methods: a Seq2Seq Modality Translation Model and a Hierarchical Seq2Seq Modality Translation Model. We also explore multiple different variations on the multimodal inputs and outputs of these seq2seq models. Our experiments on multimodal sentiment analysis using the CMU-MOSI dataset indicate that our methods learn informative multimodal representations that outperform the baselines and achieve improved performance on multimodal sentiment analysis, specifically in the Bimodal case where our model is able to improve F1 Score by twelve points. We also discuss future directions for multimodal Seq2Seq methods. | Seq2Seq2Sentiment: Multimodal Sequence to Sequence Models for Sentiment Analysis |
d15971789 | Coreference evaluation metrics are hard to optimize directly as they are nondifferentiable functions, not easily decomposable into elementary decisions. Consequently, most approaches optimize objectives only indirectly related to the end goal, resulting in suboptimal performance. Instead, we propose a differentiable relaxation that lends itself to gradient-based optimisation, thus bypassing the need for reinforcement learning or heuristic modification of cross-entropy. We show that by modifying the training objective of a competitive neural coreference system, we obtain a substantial gain in performance. This suggests that our approach can be regarded as a viable alternative to using reinforcement learning or more computationally expensive imitation learning. | Optimizing Differentiable Relaxations of Coreference Evaluation Metrics |
d202769823 | Recurrent neural networks (RNN) used for Chinese named entity recognition (NER) that sequentially track character and word information have achieved great success. However, the characteristic of chain structure and the lack of global semantics determine that RNN-based models are vulnerable to word ambiguities. In this work, we try to alleviate this problem by introducing a lexicon-based graph neural network with global semantics, in which lexicon knowledge is used to connect characters to capture the local composition, while a global relay node can capture global sentence semantics and longrange dependency. Based on the multiple graph-based interactions among characters, potential words, and the whole-sentence semantics, word ambiguities can be effectively tackled. Experiments on four NER datasets show that the proposed model achieves significant improvements against other baseline models. | A Lexicon-Based Graph Neural Network for Chinese NER |
d53081403 | Self-attention networks have proven to be of profound value for its strength of capturing global dependencies. In this work, we propose to model localness for self-attention networks, which enhances the ability of capturing useful local context. We cast localness modeling as a learnable Gaussian bias, which indicates the central and scope of the local region to be paid more attention. The bias is then incorporated into the original attention distribution to form a revised distribution. To maintain the strength of capturing long distance dependencies and enhance the ability of capturing shortrange dependencies, we only apply localness modeling to lower layers of self-attention networks. Quantitative and qualitative analyses on Chinese⇒English and English⇒German translation tasks demonstrate the effectiveness and universality of the proposed approach. | Modeling Localness for Self-Attention Networks |
d53096901 | Recent deep learning models have shown improving results to natural language generation (NLG) irrespective of providing sufficient annotated data. However, a modest training data may harm such models' performance. Thus, how to build a generator that can utilize as much of knowledge from a low-resource setting data is a crucial issue in NLG. This paper presents a variational neural-based generation model to tackle the NLG problem of having limited labeled dataset, in which we integrate a variational inference into an encoder-decoder generator and introduce a novel auxiliary autoencoding with an effective training procedure. Experiments showed that the proposed methods not only outperform the previous models when having sufficient training dataset but also show strong ability to work acceptably well when the training data is scarce.Related WorkRecently, the RNN-based generators have shown improving results in tackling the NLG problems in task oriented-dialogue systems with varied proposed methods, such as HLSTM(Wen et al., 2015a), SCLSTM (Wen et al., 2015b), or espe- | Dual Latent Variable Model for Low-Resource Natural Language Generation in Dialogue Systems |
d247158698 | Contrastive learning is emerging as a powerful technique for extracting knowledge from unlabeled data. This technique requires a balanced mixture of two ingredients: positive (similar) and negative (dissimilar) samples. This is typically achieved by maintaining a queue of negative samples during training. Prior works in the area typically uses a fixed-length negative sample queue, but how the negative sample size affects the model performance remains unclear. The opaque impact of the number of negative samples on performance when employing contrastive learning aroused our in-depth exploration. This paper presents a momentum contrastive learning model with negative sample queue for sentence embedding, namely MoCoSE. We add the prediction layer to the online branch to make the model asymmetric and together with EMA update mechanism of the target branch to prevent the model from collapsing. We define a maximum traceable distance metric, through which we learn to what extent the text contrastive learning benefits from the historical information of negative samples. Our experiments find that the best results are obtained when the maximum traceable distance is at a certain range, demonstrating that there is an optimal range of historical information for a negative sample queue. We evaluate the proposed unsupervised MoCoSE on the semantic text similarity (STS) task and obtain an average Spearman's correlation of 77.27%. Source code is available here. | Exploring the Impact of Negative Samples of Contrastive Learning: A Case Study of Sentence Embedding |
d13499279 | We study the role of the second language in bilingual word embeddings in monolingual semantic evaluation tasks. We find strongly and weakly positive correlations between down-stream task performance and second language similarity to the target language. Additionally, we show how bilingual word embeddings can be employed for the task of semantic language classification and that joint semantic spaces vary in meaningful ways across second languages. Our results support the hypothesis that semantic language similarity is influenced by both structural similarity as well as geography/contact. | Language classification from bilingual word embedding graphs |
d5678991 | We propose in this paper a new contribution to the evaluation of linguistic difficulty. At the opposite of classical approaches relying on syntax, we show that a probabilistic morpho-syntactic analysis provides information enough to calculate different parameters, including surprisal. This method constitutes an original and robust model to linguistic complexity. It constitutes a solution towards complexity experiments using spoken languages. | Predicting Linguistic Difficulty by Means of a Morpho-Syntactic Probabilistic Model |
d235097460 | Natural language processing (NLP) applications are now more powerful and ubiquitous than ever before. With rapidly developing (neural) models and ever-more available data, current NLP models have access to more information than any human speaker during their life. Still, it would be hard to argue that NLP models have reached human-level capacity. In this position paper, we argue that the reason for the current limitations is a focus on information content while ignoring language's social factors. We show that current NLP systems systematically break down when faced with interpreting the social factors of language. This limits applications to a subset of information-related tasks and prevents NLP from reaching human-level performance. At the same time, systems that incorporate even a minimum of social factors already show remarkable improvements. We formalize a taxonomy of seven social factors based on linguistic theory and exemplify current failures and emerging successes for each of them. We suggest that the NLP community address social factors to get closer to the goal of humanlike language understanding. | The Importance of Modeling Social Factors of Language: Theory and Practice |
d259106728 | RIVETER provides a complete easy-to-use pipeline for analyzing verb connotations associated with entities in text corpora. We prepopulate the package with connotation frames of sentiment, power, and agency, which have demonstrated usefulness for capturing social phenomena, such as gender bias, in a broad range of corpora. For decades, lexical frameworks have been foundational tools in computational social science, digital humanities, and natural language processing, facilitating multifaceted analysis of text corpora. But working with verb-centric lexica specifically requires natural language processing skills, reducing their accessibility to other researchers. By organizing the language processing pipeline, providing complete lexicon scores and visualizations for all entities in a corpus, and providing functionality for users to target specific research questions, RIVETER greatly improves the accessibility of verb lexica and can facilitate a broad range of future research. | RIVETER Measuring Power and Social Dynamics Between Entities |
d256627195 | We propose two methods to make unsupervised domain adaptation (UDA) more parameter efficient using adapters, small bottleneck layers interspersed with every layer of the largescale pre-trained language model (PLM). The first method deconstructs UDA into a two-step process: first by adding a domain adapter to learn domain-invariant information and then by adding a task adapter that uses domaininvariant information to learn task representations in the source domain. The second method jointly learns a supervised classifier while reducing the divergence measure. Compared to strong baselines, our simple methods perform well in natural language inference (MNLI) and the cross-domain sentiment classification task. We even outperform unsupervised domain adaptation methods such as DANN (Ganin et al., 2016) and DSN (Bousmalis et al., 2016) in sentiment classification, and we are within 0.85% F1 for natural language inference task, by fine-tuning only a fraction of the full model parameters. We release our code at https://github.com/declare-lab/domadapter. | UDAPTER -Efficient Domain Adaptation Using Adapters |
d259370543 | Non-compositional expressions present a substantial challenge for natural language processing (NLP) systems, necessitating more intricate processing compared to general language tasks, even with large pre-trained language models. Their non-compositional nature and limited availability of data resources further compound the difficulties in accurately learning their representations. This paper addresses both of these challenges. By leveraging contrastive learning techniques to build improved representations it tackles the non-compositionality challenge. Additionally, we propose a dynamic curriculum learning framework specifically designed to take advantage of the scarce available data for modeling non-compositionality. Our framework employs an easy-to-hard learning strategy, progressively optimizing the model's performance by effectively utilizing available training data. Moreover, we integrate contrastive learning into the curriculum learning approach to maximize its benefits. Experimental results demonstrate the gradual improvement in the model's performance on idiom usage recognition and metaphor detection tasks. Our evaluation encompasses six datasets, consistently affirming the effectiveness of the proposed framework. Our models available at https: //github.com/zhjjn/CLCL.git. | CLCL: Non-compositional Expression Detection with Contrastive Learning and Curriculum Learning |
d259370765 | An open-source DeepPavlov Dream Platform is specifically tailored for development of complex dialog systems like Generative AI Assistants. The stack prioritizes efficiency, modularity, scalability, and extensibility with the goal to make it easier to develop complex dialog systems from scratch. It supports modular approach to implementation of conversational agents enabling their development through the choice of NLP components and conversational skills from a rich library organized into the distributions of ready-for-use multi-skill AI assistant systems. In DeepPavlov Dream, multiskill Generative AI Assistant consists of NLP components that extract features from user utterances, conversational skills that generate or retrieve a response, skill and response selectors that facilitate choice of relevant skills and the best response, as well as a conversational orchestrator that enables creation of multi-skill Generative AI Assistants scalable up to industrial grade AI assistants. The platform allows to integrate large language models into dialog pipeline, customize with prompt engineering, handle multiple prompts during the same dialog session and create simple multimodal assistants. | DeepPavlov Dream: Platform for Building Generative AI Assistants |
d259370860 | The Ninth Revision of the International Classification of Diseases (ICD-9) is a standardized coding system used to classify health conditions. It is used for billing, tracking individual patient conditions, and for epidemiology. The highly detailed and technical nature of the codes and their associated medical conditions make it difficult for humans to accurately record them. Researchers have explored the use of neural networks, particularly language models, for automated ICD-9 code assignment. However, the imbalanced distribution of ICD-9 codes leads to poor performance. One solution is to use domain knowledge to incorporate a useful prior. This paper evaluates the usefulness of the correlation bias: we hypothesize that correlations between ICD-9 codes and other medical codes could help improve language models' performance. We showed that while the correlation bias worsens the overall performance, the effect on individual class can be negative or positive. 1 Performance on classes that are more imbalanced and less correlated with other codes is more sensitive to incorporating the correlation bias. This suggests that while the correlation bias has potential to improve ICD-9 code assignment in certain cases, the applicability criteria need to be more carefully studied. | Intriguing Effect of the Correlation Prior on ICD-9 Code Assignment |
d260063038 | In this paper we investigate potential bias in fine-tuned transformer models for irony detection. Bias is defined in this research as spurious associations between word n-grams and class labels that can cause the system to rely too much on superficial cues and miss the essence of the irony. For this purpose, we looked for correlations between class labels and words that are prone to trigger irony, such as positive adjectives, intensifiers and topical nouns. Additionally, we investigate our irony model's predictions before and after manipulating the data set through irony trigger replacements. We further support these insights with stateof-the-art explainability techniques (Layer Integrated Gradients, Discretized Integrated Gradients and Layer-wise Relevance Propagation). Both approaches confirm the hypothesis that transformer models generally encode correlations between positive sentiments and ironic texts, with even higher correlations between vividly expressed sentiment and irony. Based on these insights, we implemented a number of modification strategies to enhance the robustness of our irony classifier.Related ResearchAs of late, transformer models have become an integral part of most state-of-the-art systems for irony detection. This can be done either through direct fine-tuning(Ángel González et al., 2020)or by using the contextual embeddings from a transformer model as input for a different neural classifier, like a Convolutional Neural Network (CNN) (Ahuja and 315 | A Fine Line Between Irony and Sincerity: Identifying Bias in Transformer Models for Irony Detection |
d6844988 | In this paper we propose a new method for evaluating systems that extract temporal information from text. It uses temporal closure 1 to reward relations that are equivalent but distinct. Our metric measures the overall performance of systems with a single score, making comparison between different systems straightforward. Our approach is easy to implement, intuitive, accurate, scalable and computationally inexpensive. | Temporal Evaluation |
d170585091 | Cet article propose l'introduction d'une notion de densité syntaxique permettant de caractériser la complexité d'un énoncé et au-delà d'introduire la spécification d'un gradient de grammaticalité. Un tel gradient s'avère utile dans plusieurs cas : quantification de la difficulté d'interprétation d'une phrase, gradation de la quantité d'information syntaxique contenue dans un énoncé, explication de la variabilité et la dépendances entre les domaines linguistiques, etc. Cette notion exploite la possibilité de caractérisation fine de l'information syntaxique en termes de contraintes : la densité est fonction des contraintes satisfaites par une réalisation pour une grammaire donnée. Les résultats de l'application de cette notion à quelques corpus sont analysés.This paper introduces the notion of syntactic density that makes it possible to characterize the complexity of an utterance and to specify a gradient of grammaticality. Such a gradient is useful in several cases: quantification of the difficulty of interpreting an utterance, quantification of syntactic information of an utterance, description of variability and linguistic domains interaction, etc. This notion exploits the possibility of fine syntactic characterization in terms of constraints: density if function of satisfied constraints by an utterance for a given grammar. Some results are presented and analyzed. | Densité d'information syntaxique et gradient de grammaticalité |
d226281842 | Natural language generation (NLG) is a critical component in conversational systems, owing to its role of formulating a correct and natural text response. Traditionally, NLG components have been deployed using template-based solutions. Although neural network solutions recently developed in the research community have been shown to provide several benefits, deployment of such model-based solutions has been challenging due to high latency, correctness issues, and high data needs. In this paper, we present approaches that have helped us deploy data-efficient neural solutions for NLG in conversational systems to production. We describe a family of sampling and modeling techniques to attain production quality with light-weight neural network models using only a fraction of the data that would be necessary otherwise, and show a thorough comparison between each. Our results show that domain complexity dictates the appropriate approach to achieve high data efficiency. Finally, we distill the lessons from our experimental findings into a list of best practices for production-level NLG model development, and present them in a brief runbook. Importantly, the end products of all of the techniques are small sequence-to-sequence models (~2Mb) that we can reliably deploy in production. | Best Practices for Data-Efficient Modeling in NLG: How to Train Production-Ready Neural Models with Less Data |
d248228121 | We propose a new task for assessing machines' ability to understand fictional characters in narrative stories. The task, TVSHOWGUESS, builds on the scripts of TV series and takes the form of guessing the anonymous main characters based on the backgrounds of the scenes and dialogues. Our human study supports that this form of task covers comprehension of multiple types of character persona, including understanding characters' personalities, facts and memories of personal experience, which are well aligned with the psychological and literary theories about the theory of mind (ToM) of human beings on understanding fictional characters during reading. We further propose new model architectures to support the contextualized encoding of long scene texts. Experiments show that our proposed approaches significantly outperform baselines, yet still largely lag behind the (nearly perfect) human performance. Our work serves as a first step toward the goal of narrative character comprehension. 1 | TVSHOWGUESS: Character Comprehension in Stories as Speaker Guessing |
d259108325 | Semantic matching is a mainstream paradigm of zero-shot relation extraction, which matches a given input with a corresponding label description. The entities in the input should exactly match their hypernyms in the description, while the irrelevant contexts should be ignored when matching. However, general matching methods lack explicit modeling of the above matching pattern. In this work, we propose a fine-grained semantic matching method tailored for zero-shot relation extraction. Following the above matching pattern, we decompose the sentence-level similarity score into entity and context matching scores. Due to the lack of explicit annotations of the redundant components, we design a feature distillation module to adaptively identify the relationirrelevant features and reduce their negative impact on context matching. Experimental results show that our method achieves higher matching F 1 score and has an inference speed 10 times faster, when compared with the stateof-the-art methods. | RE-Matching: A Fine-Grained Semantic Matching Method for Zero-Shot Relation Extraction |
d441563 | We present a novel attention-based recurrent neural network for joint extraction of entity mentions and relations. We show that attention along with long short term memory (LSTM) network can extract semantic relations between entity mentions without having access to dependency trees. Experiments on Automatic Content Extraction (ACE) corpora show that our model significantly outperforms featurebased joint model byLi and Ji (2014). We also compare our model with an end-toend tree-based LSTM model (SPTree) byMiwa and Bansal (2016)and show that our model performs within 1% on entity mentions and 2% on relations. Our finegrained analysis also shows that our model performs significantly better on AGENT-ARTIFACT relations, while SPTree performs better on PHYSICAL and PART-WHOLE relations. | Going out on a limb: Joint Extraction of Entity Mentions and Relations without Dependency Trees |
d216080642 | Organisations disclose their privacy practices by posting privacy policies on their websites. Even though internet users often care about their digital privacy, they usually do not read privacy policies, since understanding them requires a significant investment of time and effort. Natural language processing has been used to create experimental tools to interpret privacy policies, but there has been a lack of large privacy policy corpora to facilitate the creation of large-scale semi-supervised and unsupervised models to interpret and simplify privacy policies. Thus, we present the Pri-vaSeer Corpus of 1,005,380 English language website privacy policies collected from the web. The number of unique websites represented in PrivaSeer is about ten times larger than the next largest public collection of web privacy policies, and it surpasses the aggregate of unique websites represented in all other publicly available privacy policy corpora combined. We describe a corpus creation pipeline with stages that include a web crawler, language detection, document classification, duplicate and near-duplicate removal, and content extraction. We employ an unsupervised topic modelling approach to investigate the contents of policy documents in the corpus and discuss the distribution of topics in privacy policies at web scale. We further investigate the relationship between privacy policy domain PageRanks and text features of the privacy policies. Finally, we use the corpus to pretrain PrivBERT, a transformer-based privacy policy language model, and obtain state of the art results on the data practice classification and question answering tasks. | Privacy at Scale: Introducing the PrivaSeer Corpus of Web Privacy Policies |
d247939526 | Existing reference-free metrics have obvious limitations for evaluating controlled text generation models. Unsupervised metrics can only provide a task-agnostic evaluation result which correlates weakly with human judgments, whereas supervised ones may overfit task-specific data with poor generalization ability to other datasets. In this paper, we propose an unsupervised reference-free metric called CTRLEval, which evaluates controlled text generation from different aspects by formulating each aspect into multiple text infilling tasks. On top of these tasks, the metric assembles the generation probabilities from a pre-trained language model without any model training. Experimental results show that our metric has higher correlations with human judgments than other baselines, while obtaining better generalization of evaluating generated texts from different models and with different qualities 1 . | CTRLEval: An Unsupervised Reference-Free Metric for Evaluating Controlled Text Generation |
d8169368 | There is a growing interest in researching null instantiations, which are those implicit semantic arguments. Many of these implicit arguments can be linked to referents in context, and their discoveries are of great benefits to semantic processing. We address the issue of automatically identifying and resolving implicit arguments in Chinese discourse. For their resolutions, we present an approach that combines the information about overtly labeled arguments and frame-to-frame relations defined by FrameNet. Experimental results on our created corpus demonstrate the effectiveness of our approach. | Implicit Role Linking on Chinese Discourse: Exploiting Explicit Roles and Frame-to-Frame Relations |
d8532241 | Nous proposons ici des méthodes de désambiguisation sémantique par substition lexicale pour la tâche 1 de l'atelier SemDis2014. Les méthodes exposées dans ce papier sont toutes bâties à partir de balades aléatoires courtes dans des graphes unipartis ou bipartis construits sur diverses ressources. Certaines de ces méthodes n'utilisent que des graphes construits automatiquement à partir de corpus (méthodes non supervisées), d'autres utilisent des graphes construits à partir de ressources produites « à la main » par des lexicographes ou par les foules (méthodes supervisées).Abstract. In this paper, we propose word sense disambiguation methods based on lexical substitution and used for the task 1 of the SemDis2014 workshop. This methods are run by using short random walks on unipartite networks or bipartite networks. Some of these methods only use graphs automatically built from corpora (unsurpervised methods), others also use graphs built from handcraft resources filled by lexicographers or by the crowds (supervised methods).Mots-clés : désambiguisation sémantique, substition lexicale, réseaux lexicaux, balades aléatoires courtes.Keywords: word sense disambiguation, lexical substitution, lexical networks, short random walks.2. Méthode qui s'appuie sur des ressources lexicales de type dictionnairique. 3. Méthode de catégorie (b). 4. Cette dénomination peut cependant être abusive dans la mesure où le système automatique pourrait éventuellement utiliser dans sa chaîne de traitements des ressources construites « à la main », par exemple quand la chaîne de traitements utilise un analyseur syntaxique qui lui même utilise des ressources construites « à la main ».5. C'est la phrase numéro 93 du jeu de test fourni par SemDis2014. | 21 ème Traitement Automatique des Langues Naturelles |
d248887569 | Large pretrained multilingual models, trained on dozens of languages, have delivered promising results due to cross-lingual learning capabilities on a variety of language tasks. Further adapting these models to specific languages, especially ones unseen during pre-training, is an important goal toward expanding the coverage of language technologies. In this study, we show how we can use language phylogenetic information to improve cross-lingual transfer leveraging closely related languages in a structured, linguistically-informed manner. We perform adapter-based training on languages from diverse language families (Germanic, Uralic, Tupian, Uto-Aztecan) and evaluate on both syntactic and semantic tasks, obtaining more than 20% relative performance improvements over strong commonly used baselines, especially on languages unseen during pre-training. 1 Timothy Dozat and Christopher D. Manning. 2017.Deep biaffine attention for neural dependency pars- | Phylogeny-Inspired Adaptation of Multilingual Models to New Languages |
d20090034 | Current word alignment models do not distinguish between different types of alignment links. In this paper, we provide a new probabilistic model for word alignment where word alignments are associated with linguistically motivated alignment types. We propose a novel task of joint prediction of word alignment and alignment types and propose novel semi-supervised learning algorithms for this task. We also solve a sub-task of predicting the alignment type given an aligned word pair. In our experimental results, the generative models we introduce to model alignment types significantly outperform the models without alignment types. | Joint Prediction of Word Alignment with Alignment Types |
d253254835 | Pre-trained Transformers currently dominate most NLP tasks. They impose, however, limits on the maximum input length (512 subwords in BERT), which are too restrictive in the legal domain. Even sparse-attention models, such as Longformer and BigBird, which increase the maximum input length to 4,096 sub-words, severely truncate texts in three of the six datasets of LexGLUE. Simpler linear classifiers with TF-IDF features can handle texts of any length, require far less resources to train and deploy, but are usually outperformed by pre-trained Transformers. We explore two directions to cope with long legal texts: (i) modifying a Longformer warm-started from LegalBERT to handle even longer texts (up to 8,192 sub-words), and (ii) modifying Legal-BERT to use TF-IDF representations. The first approach is the best in terms of performance, surpassing a hierarchical version of LegalBERT, which was the previous state of the art in LexGLUE. The second approach leads to computationally more efficient models at the expense of lower performance, but the resulting models still outperform overall a linear SVM with TF-IDF features in long legal document classification. | Processing Long Legal Documents with Pre-trained Transformers: Modding LegalBERT and Longformer |
d258947371 | Current pre-trained language models have enabled remarkable improvements in downstream tasks, but it remains difficult to distinguish effects of statistical correlation from more systematic logical reasoning grounded on the understanding of real world. We tease these factors apart by leveraging counterfactual conditionals, which force language models to predict unusual consequences based on hypothetical propositions. We introduce a set of tests from psycholinguistic experiments, as well as larger-scale controlled datasets, to probe counterfactual predictions from five pre-trained language models. We find that models are consistently able to override real-world knowledge in counterfactual scenarios, and that this effect is more robust in case of stronger baseline world knowledge-however, we also find that for most models this effect appears largely to be driven by simple lexical cues. When we mitigate effects of both world knowledge and lexical cues to test knowledge of linguistic nuances of counterfactuals, we find that only GPT-3 shows sensitivity to these nuances, though this sensitivity is also non-trivially impacted by lexical associative factors. 1 | Counterfactual reasoning: Testing language models' understanding of hypothetical scenarios |
d248085885 | Media news framing bias can increase political polarization and undermine civil society. The need for automatic mitigation methods is therefore growing. We propose a new task, a neutral summary generation from multiple news articles of the varying political leanings to facilitate balanced and unbiased news reading. In this paper, we first collect a new dataset, illustrate insights about framing bias through a case study, and propose a new effective metric and model (NEUS-TITLE) for the task. Based on our discovery that title provides a good signal for framing bias, we present NEUS-TITLE that learns to neutralize news content in hierarchical order from title to article. Our hierarchical multi-task learning is achieved by formatting our hierarchical data pair (title, article) sequentially with identifier-tokens ("TI-TLE=>", "ARTICLE=>") and fine-tuning the auto-regressive decoder with the standard negative log-likelihood objective. We then analyze and point out the remaining challenges and future directions. One of the most interesting observations is that neural NLG models can hallucinate not only factually inaccurate or unverifiable content but also politically biased content. | NeuS: Neutral Multi-News Summarization for Mitigating Framing Bias |
d51812449 | Natural Language Inference (NLI), also known as Recognizing Textual Entailment (RTE), is one of the most important problems in natural language processing. It requires to infer the logical relationship between two given sentences. While current approaches mostly focus on the interaction architectures of the sentences, in this paper, we propose to transfer knowledge from some important discourse markers to augment the quality of the NLI model. We observe that people usually use some discourse markers such as "so" or "but" to represent the logical relationship between two sentences. These words potentially have deep connections with the meanings of the sentences, thus can be utilized to help improve the representations of them. Moreover, we use reinforcement learning to optimize a new objective function with a reward defined by the property of the NLI datasets to make full use of the labels information. Experiments show that our method achieves the state-of-the-art performance on several large-scale datasets. 1 Here sentences mean either the whole sentences or the main clauses of a compound sentence. | Discourse Marker Augmented Network with Reinforcement Learning for Natural Language Inference |
d298145 | Most of previous work in knowledge base (KB) completion has focused on the problem of relation extraction. In this work, we focus on the task of inferring missing entity type instances in a KB, a fundamental task for KB competition yet receives little attention.Due to the novelty of this task, we construct a large-scale dataset and design an automatic evaluation methodology. Our knowledge base completion method uses information within the existing KB and external information from Wikipedia. We show that individual methods trained with a global objective that considers unobserved cells from both the entity and the type side gives consistently higher quality predictions compared to baseline methods. We also perform manual evaluation on a small subset of the data to verify the effectiveness of our knowledge base completion methods and the correctness of our proposed automatic evaluation method. | Inferring Missing Entity Type Instances for Knowledge Base Completion: New Dataset and Methods |
d15215411 | Plagiarism, the unacknowledged reuse of text, does not end at language boundaries. Cross-language plagiarism occurs if a text is translated from a fragment written in a different language and no proper citation is provided. Regardless of the change of language, the contents and, in particular, the ideas remain the same. Whereas different methods for the detection of monolingual plagiarism have been developed, less attention has been paid to the crosslanguage case.In this paper we compare two recently proposed cross-language plagiarism detection methods (CL-CNG, based on character n-grams and CL-ASA, based on statistical translation), to a novel approach to this problem, based on machine translation and monolingual similarity analysis (T+MA). We explore the effectiveness of the three approaches for less related languages. CL-CNG shows not be appropriate for this kind of language pairs, whereas T+MA performs better than the previously proposed models. | Plagiarism Detection across Distant Language Pairs |
d15295411 | Variation in language is ubiquitous, particularly in newer forms of writing such as social media. Fortunately, variation is not random; it is often linked to social properties of the author. In this paper, we show how to exploit social networks to make sentiment analysis more robust to social language variation. The key idea is linguistic homophily: the tendency of socially linked individuals to use language in similar ways. We formalize this idea in a novel attention-based neural network architecture, in which attention is divided among several basis models, depending on the author's position in the social network. This has the effect of smoothing the classification function across the social network, and makes it possible to induce personalized classifiers even for authors for whom there is no labeled data or demographic metadata. This model significantly improves the accuracies of sentiment analysis on Twitter and on review data.We propose a middle ground between group-level demographic characteristics and personalization, by exploiting social network structure. The sociological theory of homophily asserts that individuals are usually similar to their friends(McPherson et al., 2001). This property has been demonstrated for language(Bryden et al., 2013)as well as for the demographic properties targeted by Hovy(2015), which are more likely to be shared by friends than by random pairs of individuals(Thelwall, 2009). Social | Overcoming Language Variation in Sentiment Analysis with Social Attention |
d234341137 | Very recently, few certified defense methods have been developed to provably guarantee the robustness of a text classifier to adversarial synonym substitutions. However, all the existing certified defense methods assume that the defenders have been informed of how the adversaries generate synonyms, which is not a realistic scenario. In this study, we propose a certifiably robust defense method by randomly masking a certain proportion of the words in an input text, in which the above unrealistic assumption is no longer necessary. The proposed method can defend against not only word substitution-based attacks, but also character-level perturbations. We can certify the classifications of over 50% of texts to be robust to any perturbation of five words on AGNEWS, and two words on SST2 dataset. The experimental results show that our randomized smoothing method significantly outperforms recently proposed defense methods across multiple datasets under different attack algorithms. | Certified Robustness to Text Adversarial Attacks by Randomized [MASK] under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license |
d38866101 | JeuxDeMots (JDM), Lafourcade(2007)is a French lexical-semantic network. It is a knowledge base containing lexical and semantic information. The network is composed of terms (nodes) and relations (edges). The relations between nodes are typed, oriented and weighted. Around 100 relation types are defined, such as synonymy, antonymy, generic (hypernymy), specific (hyponymy) and refinements. Refinements are representations of word senses or usages.The different refinements of a given term T take the form of (T, glosses) pairs, as T>glose 1 , T>glose 2 , ..., T>glose n . Glosses are terms that help the reader to identify the proper meaning of T. For instance, the French term frégate (frigate), which is a ship and a bird, has two refinements, frégate>navire and frégate>oiseau. Thus, a term T is linked to its refinements in the network, through a specific relation type (r_semantic_raff ). | Identifying Polysemous Words and Inferring Sense Glosses in a Semantic Network |
d248478191 | This paper aims for a potential architectural improvement for multilingual learning and asks: Can different tasks from different languages be modeled in a monolithic framework, i.e. without any task/language-specific module? The benefit of achieving this could open new doors for future multilingual research, including allowing systems trained on low resources to be further assisted by other languages as well as other tasks. We approach this goal by developing a learning framework named Polyglot Prompting to exploit prompting methods for learning a unified semantic space for different languages and tasks with multilingual prompt engineering. We performed a comprehensive evaluation of 6 tasks, namely topic classification, sentiment classification, named entity recognition, question answering, natural language inference, and summarization, covering 24 datasets and 49 languages. The experimental results demonstrated the efficacy of multilingual multitask prompt-based learning and led to inspiring observations. We also present an interpretable multilingual evaluation methodology and show how the proposed framework, multilingual multitask prompt training, works. We release all datasets prompted in the best setting and code. | Polyglot Prompting: Multilingual Multitask Prompt Training |
d259370527 | Writing assistants are valuable tools that can help writers improve their writing skills. We introduce Effidit (Efficient and Intelligent Editing), a digital writing assistant that facilitates users to write higher-quality text more efficiently through the use of Artificial Intelligence (AI) and Natural Language Processing (NLP) technologies. We significantly expand the capacities of a writing assistant by providing functions in three modules: text completion, hint recommendation, and writing refinement. Based on the above efforts, Effidit can efficiently assist users in creating their own text. Effidit has been deployed to several Tencent products and publicly released at https://effidit.qq.com/. | Effidit: An Assistant for Improving Writing Efficiency |
d259370579 | In work on AMR (Abstract Meaning Representation), similarity metrics are crucial as they are used to evaluate AMR systems such as AMR parsers. Current AMR metrics are all based on nodes or triples matching without considering the entire structures of AMR graphs. To address this problem, and inspired by learned similarity evaluation on plain text, we propose AMRSim, an automatic AMR graph similarity evaluation metric. To overcome the high cost of collecting human-annotated data, AMRSim automatically generates silver AMR graphs and utilizes self-supervised learning methods. We evaluated AMRSim on various datasets and found that AMRSim significantly improves the correlations with human semantic scores and remains robust under diverse challenges. We also discuss how AMRSim can be extended to multilingual cases. 1 | Evaluate AMR Graph Similarity via Self-supervised Learning |
d259370620 | We revisit the multimodal entity and relation extraction from a translation point of view. Special attention is paid on the misalignment issue in text-image datasets which may mislead the learning. We are motivated by the fact that the cross-modal misalignment is a similar problem of cross-lingual divergence issue in machine translation. The problem can then be transformed and existing solutions can be borrowed by treating a text and its paired image as the translation to each other. We implement a multimodal back-translation using diffusionbased generative models for pseudo-paralleled pairs and a divergence estimator by constructing a high-resource corpora as a bridge for low-resource learners. Fine-grained confidence scores are generated to indicate both types and degrees of alignments with which better representations are obtained. | Rethinking Multimodal Entity and Relation Extraction from a Translation Point of View |
d246863418 | Effectively scaling large Transformer models is a main driver of recent advances in natural language processing. Dynamic neural networks, as an emerging research direction, are capable of scaling up neural networks with sub-linear increases in computation and time by dynamically adjusting their computational path based on the input. Dynamic neural networks could be a promising solution to the growing parameter numbers of pretrained language models, allowing both model pretraining with trillions of parameters and faster inference on mobile devices. In this survey, we summarize the progress of three types of dynamic neural networks in NLP: skimming, mixture of experts, and early exit. We also highlight current challenges in dynamic neural networks and directions for future research. | A Survey on Dynamic Neural Networks for Natural Language Processing |
d226278305 | Answer validation in machine reading comprehension (MRC) consists of verifying an extracted answer against an input context and question pair. Previous work has looked at re-assessing the "answerability" of the question given the extracted answer. Here we address a different problem: the tendency of existing MRC systems to produce partially correct answers when presented with answerable questions. We explore the nature of such errors and propose a post-processing correction method that yields statistically significant performance improvements over state-of-the-art MRC systems in both monolingual and multilingual evaluation. | Answer Span Correction in Machine Reading Comprehension |
d235266206 | Unlike English letters, Chinese characters have rich and specific meanings. Usually, the meaning of a word can be derived from its constituent characters in some way. Several previous works on syntactic parsing propose to annotate shallow word-internal structures for better utilizing character-level information. This work proposes to model the deep internal structures of Chinese words as dependency trees with 11 labels for distinguishing syntactic relationships. First, based on newly compiled annotation guidelines, we manually annotate a word-internal structure treebank (WIST) consisting of over 30K multi-char words from Chinese Penn Treebank. To guarantee quality, each word is independently annotated by two annotators and inconsistencies are handled by a third senior annotator. Second, we present detailed and interesting analysis on WIST to reveal insights on Chinese word formation. Third, we propose word-internal structure parsing as a new task, and conduct benchmark experiments using a competitive dependency parser. Finally, we present two simple ways to encode word-internal structures, leading to promising gains on the sentence-level syntactic parsing task. | An In-depth Study on Internal Structure of Chinese Words |
d215737184 | We present the Neural Covidex, a search engine that exploits the latest neural ranking architectures to provide information access to the COVID-19 Open Research Dataset curated by the Allen Institute for AI. This web application exists as part of a suite of tools that we have developed over the past few weeks to help domain experts tackle the ongoing global pandemic. We hope that improved information access capabilities to the scientific literature can inform evidence-based decision making and insight generation. This paper describes our initial efforts and offers a few thoughts about lessons we have learned along the way. | |
d2718048 | Basilica is an event-driven software architecture for creating conversational agents as a collection of reusable components. Software engineers and computer scientists can use this general architecture to create increasingly sophisticated conversational agents. We have developed agents based on Basilica that have been used in various application scenarios and foresee that agents build on Basilica can cater to a wider variety of interactive situations as we continue to add functionality to our architecture. | Building Conversational Agents with Basilica |
d248721755 | Task transfer, transferring knowledge contained in related tasks, holds the promise of reducing the quantity of labeled data required to finetune language models. Dialogue understanding encompasses many diverse tasks, yet task transfer has not been thoroughly studied in conversational AI. This work explores conversational task transfer by introducing FETA: a benchmark for FEw-sample TAsk transfer in open-domain dialogue. FETA contains two underlying sets of conversations upon which there are 10 and 7 tasks annotated, enabling the study of intra-dataset task transfer; task transfer without domain adaptation. We utilize three popular language models and three learning algorithms to analyze the transferability between 132 source-target task pairs and create a baseline for future work. We run experiments in the single-and multi-source settings and report valuable findings, e.g., most performance trends are model-specific, and span extraction and multiple-choice tasks benefit the most from task transfer. In addition to task transfer, FETA can be a valuable resource for future research into the efficiency and generalizability of pre-training datasets and model architectures, as well as for learning settings such as continual and multitask learning. | FETA: A Benchmark for Few-Sample Task Transfer in Open-Domain Dialogue |
d258378152 | Machine Translation operates on the premise of an interlingua which abstracts away from the surface form while preserving the meaning. A decade ago, the idea of using round-trip MT to guide Grammatical Error Correction was proposed as a way to abstract away from potential errors in surface forms(Madnani et al., 2012). At the time, it did not pan out due to the low quality of MT systems of the day. Today much stronger MT systems are available so we re-evaluate this idea across five languages and models of various sizes. We find that for extra large models input augmentation through round-trip MT has little to no effect. For more 'workable' model sizes, however, it yields consistent improvements, sometimes bringing the performance of a base or large model up to that of a large or xl model, respectively. The round-trip translation comes at a computational cost though, so one would have to determine whether to opt for a larger model or for input augmentation on a case-by-case basis. | Grammatical Error Correction through Round-Trip Machine Translation |
d258378275 | Contrastive Language-Image Pre-training (CLIP) has shown remarkable success in learning with cross-modal supervision from extensive amounts of image-text pairs collected online. Thus far, the effectiveness of CLIP has been investigated primarily in general-domain multimodal problems. In this work, we evaluate the effectiveness of CLIP for the task of Medical Visual Question Answering (MedVQA). We present PubMedCLIP, a fine-tuned version of CLIP for the medical domain based on PubMed articles. Our experiments conducted on two MedVQA benchmark datasets illustrate that PubMed-CLIP achieves superior results improving the overall accuracy up to 3% in comparison to the state-of-the-art Model-Agnostic Meta-Learning (MAML) networks pre-trained only on visual data. The PubMedCLIP model with different back-ends, the source code for pre-training them and reproducing our MedVQA pipeline is publicly available at https://github.com/sarahESL/PubMedCLIP. | PubMedCLIP: How Much Does CLIP Benefit Visual Question Answering in the Medical Domain? |
d258486963 | On the Nature of Discrete Speech Representations in Multilingual Self-supervised Models | |
d245425218 | Tokenization is fundamental to pretrained language models (PLMs). Existing tokenization methods for Chinese PLMs typically treat each character as an indivisible token. However, they ignore the unique feature of the Chinese writing system where additional linguistic information exists below the character level, i.e., at the sub-character level. To utilize such information, we propose sub-character (SubChar for short) tokenization. Specifically, we first encode the input text by converting each Chinese character into a short sequence based on its glyph or pronunciation, and then construct the vocabulary based on the encoded text with sub-word segmentation. Experimental results show that SubChar tokenizers have two main advantages over existing tokenizers: 1) They can tokenize inputs into much shorter sequences, thus improving the computational efficiency. 2) Pronunciation-based SubChar tokenizers can encode Chinese homophones into the same transliteration sequences and produce the same tokenization output, hence being robust to homophone typos. At the same time, models trained with SubChar tokenizers perform competitively on downstream tasks. We release our code and models at https://github.com/thunlp /SubCharTokenization to facilitate future work.469 | Sub-Character Tokenization for Chinese Pretrained Language Models |
d259376506 | This paper presents D2KLab's system used for the shared task of "Multilingual Complex Named Entity Recognition (MultiCoNER II)", as part of SemEval 2023 Task 2. The system relies on a fine-tuned transformer based language model for extracting named entities. We present the architecture of the system, and we discuss our results and observations. Our implementation is open sourced at https://github.com/D2KLab/multiconer. | D2KLab at SemEval-2023 Task 2: Leveraging T-NER to Develop a Fine-Tuned Multilingual Model for Complex Named Entity Recognition |
d259376706 | We describe our participation on the Multievidence Natural Language Inference for Clinical Trial Data (NLI4CT) of SemEval'23. The organizers provided a collection of clinical trials as training data and a set of statements, which can be related to either a single trial or to a comparison of two trials. The task consisted of two sub-tasks: (i) textual entailment (Task 1) for predicting whether the statement is supported (Entailment) or not (Contradiction) by the corresponding trial(s); and (ii) evidence retrieval (Task 2) for selecting the evidences (sentences in the trials) that support the decision made for Task 1. We built a model based on a sentence-based BERT similarity model which was pre-trained on ClinicalBERT embeddings. Our best results on the official test sets were f-scores of 0.64 and 0.67 for Tasks 1 and 2, respectively. | Bf3R at SemEval-2023 Task 7: a text similarity model for textual entailment and evidence retrieval in clinical trials and animal studies |
d18357549 | The aim of the Helping Our Own (HOO) Shared Task is to promote the development of automated tools and techniques that can assist authors in the writing task, with a specific focus on writing within the natural language processing community. This paper reports on the results of a pilot run of the shared task, in which six teams participated. We describe the nature of the task and the data used, report on the results achieved, and discuss some of the things we learned that will guide future versions of the task. | Helping Our Own: The HOO 2011 Pilot Shared Task |
d253510351 | Event argument extraction has long been studied as a sequential prediction problem with extractive-based methods, tackling each argument in isolation. Although recent work proposes generation-based methods to capture cross-argument dependency, they require generating and post-processing a complicated target sequence (template). Motivated by these observations and recent pretrained language models' capabilities of learning from demonstrations. We propose a retrieval-augmented generative QA model (R-GQA) for event argument extraction. It retrieves the most similar QA pair and augments it as prompt to the current example's context, then decodes the arguments as answers. Our approach outperforms substantially prior methods across various settings (i.e. fully supervised, domain transfer, and fewshot learning). Finally, we propose a clusteringbased sampling strategy (JointEnc) and conduct a thorough analysis of how different strategies influence the few-shot learning performance. 1 | Retrieval-Augmented Generative Question Answering for Event Argument Extraction |
d15854992 | Today's natural language processing systems are growing more complex with the need to incorporate a wider range of language resources and more sophisticated statistical methods. In many cases, it is necessary to learn a component with input that includes the predictions of other learned components or to assign simultaneously the values that would be assigned by multiple components with an expressive, data dependent structure among them. As a result, the design of systems with multiple learning components is inevitably quite technically complex, and implementations of conceptually simple NLP systems can be time consuming and prone to error. Our new modeling language, Learning Based Java (LBJ), facilitates the rapid development of systems that learn and perform inference. LBJ has already been used to build state of the art NLP systems. This paper details recent advancements in the language which generalize its computational model, making a wider class of algorithms available. | Learning Based Java for Rapid Development of NLP Systems |
d219963636 | Discourse representation structures (DRSs) are scoped semantic representations for texts of arbitrary length. Evaluation of the accuracy of predicted DRSs plays a key role in developing semantic parsers and improving their performance. DRSs are typically visualized as nested boxes, in a way that is not straightforward to process automatically. COUNTER, an evaluation algorithm for DRSs, transforms them to clauses and measures clause overlap by searching for variable mappings between two DRSs. Unfortunately, COUNTER is computationally costly (with respect to memory and CPU time) and does not scale with longer texts. We introduce DSCORER, an efficient new metric which converts box-style DRSs to graphs and then measures the overlap of n-grams in the graphs. Experiments show that DSCORER computes accuracy scores that correlate with scores from COUNTER at a fraction of the time. | DSCORER: A Fast Evaluation Metric for Discourse Representation Structure Parsing |
d219177032 | In this paper, we describe an approach for modelling causal reasoning in natural language by detecting counterfactuals in text using multi-head self-attention weights. We use pre-trained transformer models to extract contextual embeddings and self-attention weights from the text. We show the use of convolutional layers to extract task-specific features from these self-attention weights. Further, we describe a fine-tuning approach with a common base model for knowledge sharing between the two closely related sub-tasks for counterfactual detection. We analyze and compare the performance of various transformer models in our experiments. Finally, we perform a qualitative analysis with the multi-head self-attention weights to interpret our models' dynamics. | CNRL at SemEval-2020 Task 5: Modelling Causal Reasoning in Language with Multi-Head Self-Attention Weights based Counterfactual Detection |
d220055815 | Code-mixing is the phenomenon of using multiple languages in the same utterance of a text or speech. It is a frequently used pattern of communication on various platforms such as social media sites, online gaming, product reviews, etc. Sentiment analysis of the monolingual text is a well-studied task. Code-mixing adds to the challenge of analyzing the sentiment of the text due to the non-standard writing style. We present a candidate sentence generation and selection based approach on top of the Bi-LSTM based neural classifier to classify the Hinglish code-mixed text into one of the three sentiment classes positive, negative, or neutral. The proposed approach shows an improvement in the system performance as compared to the Bi-LSTM based neural classifier. The results present an opportunity to understand various other nuances of code-mixing in the textual data, such as humor-detection, intent classification, etc. | IIT Gandhinagar at SemEval-2020 Task 9: Code-Mixed Sentiment Classification Using Candidate Sentence Generation and Selection |
d238583377 | NLP models that compare or consolidate information across multiple documents often struggle when challenged with recognizing substantial information redundancies across the texts. For example, in multi-document summarization it is crucial to identify salient information across texts and then generate a non-redundant summary, while facing repeated and usually differently-phrased salient content. To facilitate researching such challenges, the sentencelevel task of sentence fusion was proposed, yet previous datasets for this task were very limited in their size and scope. In this paper, we revisit and substantially extend previous dataset creation efforts. With careful modifications, relabeling and employing complementing data sources, we were able to triple the size of a notable earlier dataset. Moreover, we show that our extended version uses more representative texts for multi-document tasks and provides a larger and more diverse training set, which substantially improves model training. | Extending Multi-Text Sentence Fusion Resources via Pyramid Annotations |
d16725678 | Current machine translation (MT) techniques are continuously improving. In specific areas, post-editing (PE) can enable the production of high-quality translations relatively quickly. But is it feasible to translate a literary work (fiction, short story, etc) using such an MT+PE pipeline? This paper offers an initial response to this question. An essay by the American writer Richard Powers, currently not available in French, is automatically translated and post-edited and then revised by non-professional translators. In addition to presenting experimental evaluation results of the MT+PE pipeline (MT system used, automatic evaluation), we also discuss the quality of the translation output from the perspective of a panel of readers (who read the translated short story in French, and answered a survey afterwards). Finally, some remarks of the official French translator of R. Powers, requested on this occasion, are given at the end of this article. | Automated translation of a literary work: a pilot study |
d52179536 | Responses in task-oriented dialogue systems often realize multiple propositions whose ultimate form depends on the use of sentence planning and discourse structuring operations. For example a recommendation may consist of an explicitly evaluative utterance e.g. Chanpen Thai is the best option, along with content related by the justification discourse relation, e.g. It has great food and service, that combines multiple propositions into a single phrase. While neural generation methods integrate sentence planning and surface realization in one endto-end learning framework, previous work has not shown that neural generators can:(1) perform common sentence planning and discourse structuring operations; (2) make decisions as to whether to realize content in a single sentence or over multiple sentences; (3) generalize sentence planning and discourse relation operations beyond what was seen in training. We systematically create large training corpora that exhibit particular sentence planning operations and then test neural models to see what they learn. We compare models without explicit latent variables for sentence planning with ones that provide explicit supervision during training. We show that only the models with additional supervision can reproduce sentence planning and discourse operations and generalize to situations unseen in training.# Type ExamplePRICERANGE[MODERATE], AREA[RIVERSIDE], NAME[ZIZZI], FOOD[ENGLISH], EATTYPE[PUB] NEAR[AVALON], FAMILYFRIENDLY[NO] | Can Neural Generators for Dialogue Learn Sentence Planning and Discourse Structuring? |
d17342618 | Early diagnosis of neurodegenerative disorders (ND) such as Alzheimer's disease (AD) and related Dementias is currently a challenge. Currently, AD can only be diagnosed by examining the patient's brain after death and Dementia is diagnosed typically through consensus using specific diagnostic criteria and extensive neuropsychological examinations with tools such as the Mini-Mental State Examination (MMSE) or the Montreal Cognitive Assessment (MoCA). In this paper, we use several Machine Learning (ML) algorithms to build diagnostic models using syntactic and lexical features resulting from verbal utterances of AD and related Dementia patients. We emphasize that the best diagnostic model distinguished the AD and related Dementias group from the healthy elderly group with 74% F-Measure using Support Vector Machines (SVM). Additionally, we perform several statistical tests to indicate the significance of the selected linguistic features. Our results show that syntactic and lexical features could be good indicative features for helping to diagnose AD and related Dementias. | Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal Learning Predictive Linguistic Features for Alzheimer's Disease and related Dementias using Verbal Utterances |
d18210792 | We aim at showing that lexical descriptions based on multifactorial and continuous models can be used by linguists and lexicographers (and not only by machines) so long as they are provided with a way to efficiently navigate data collections. We propose to demonstrate such a system. | Exploring a Continuous and Flexible Representation of the Lexicon |
d252819038 | ICD coding is designed to assign the disease codes to electronic health records (EHRs) upon discharge, which is crucial for billing and clinical statistics. In an attempt to improve the effectiveness and efficiency of manual coding, many methods have been proposed to automatically predict ICD codes from clinical notes. However, most previous works ignore the decisive information contained in structured medical data in EHRs, which is hard to be captured from the noisy clinical notes. In this paper, we propose a Tree-enhanced Multimodal Attention Network (TreeMAN) to fuse tabular features and textual features into multimodal representations by enhancing the text representations with tree-based features via the attention mechanism. Tree-based features are constructed according to decision trees learned from structured multimodal medical data, which capture the decisive information about ICD coding. We can apply the same multi-label classifier from previous text models to the multimodal representations to predict ICD codes. Experiments on two MIMIC datasets show that our method outperforms prior state-of-the-art ICD coding approaches. | TreeMAN: Tree-enhanced Multimodal Attention Network for ICD Coding |
d253083210 | Memes are a powerful tool for communication over social media. Their affinity for evolving across politics, history, and sociocultural phenomena makes them an ideal communication vehicle. To comprehend the subtle message conveyed within a meme, one must understand the background that facilitates its holistic assimilation. Besides digital archiving of memes and their metadata by a few websites like knowyourmeme.com, currently, there is no efficient way to deduce a meme's context dynamically. In this work, we propose a novel task, MEMEX -given a meme and a related document, the aim is to mine the context that succinctly explains the background of the meme. At first, we develop MCC (Meme Context Corpus), a novel dataset for MEMEX. Further, to benchmark MCC, we propose MIME (MultImodal Meme Explainer), a multimodal neural framework that uses common sense enriched meme representation and a layered approach to capture the cross-modal semantic dependencies between the meme and the context. MIME surpasses several unimodal and multimodal systems and yields an absolute improvement of ≈ 4% F1-score over the best baseline. Lastly, we conduct detailed analyses of MIME's performance, highlighting the aspects that could lead to optimal modeling of cross-modal contextual associations. | MEMEX: Detecting Explanatory Evidence for Memes via Knowledge-Enriched Contextualization |
d238856938 | Deep NLP models have been shown to be brittle to input perturbations. Recent work has shown that data augmentation using counterfactuals -i.e. minimally perturbed inputscan help ameliorate this weakness. We focus on the task of creating counterfactuals for question answering, which presents unique challenges related to world knowledge, semantic diversity, and answerability. To address these challenges, we develop a Retrieve-Generate-Filter (RGF) technique to create counterfactual evaluation and training data with minimal human supervision. Using an open-domain QA framework and question generation model trained on original task data, we create counterfactuals that are fluent, semantically diverse, and automatically labeled. Data augmentation with RGF counterfactuals improves performance on out-of-domain and challenging evaluation sets over and above existing methods, in both the reading comprehension and open-domain QA settings. Moreover, we find that RGF data leads to significant improvements to robustness to local perturbations. 1 | Retrieval-guided Counterfactual Generation for QA |
d9211323 | This paper evaluates four metaphor identification systems on the 200,000 word VU Amsterdam Metaphor Corpus, comparing results by genre and by sub-class of metaphor. The paper then compares the rate of agreement between the systems for each genre and sub-class. Each of the identification systems is based, explicitly or implicitly, on a theory of metaphor which hypothesizes that certain properties are essential to metaphor-inlanguage. The goal of this paper is to see what the success or failure of these systems can tell us about the essential properties of metaphorin-language. The success of the identification systems varies significantly across genres and sub-classes of metaphor. At the same time, the different systems achieve similar success rates on each even though they show low agreement among themselves. This is taken to be evidence that there are several sub-types of metaphor-in-language and that the ideal metaphor identification system will first define these sub-types and then model the linguistic properties which can distinguish these sub-types from one another and from nonmetaphors. | What metaphor identification systems can tell us about metaphor-in-language |
d256615810 | Several proposals have been put forward in recent years for improving out-of-distribution (OOD) performance through mitigating dataset biases. A popular workaround is to train a robust model by re-weighting training examples based on a secondary biased model. Here, the underlying assumption is that the biased model resorts to shortcut features. Hence, those training examples that are correctly predicted by the biased model are flagged as being biased and are down-weighted during the training of the main model. However, assessing the importance of an instance merely based on the predictions of the biased model may be too naive. It is possible that the prediction of the main model can be derived from another decisionmaking process that is distinct from the behavior of the biased model. To circumvent this, we introduce a fine-tuning strategy that incorporates the similarity between the main and biased model attribution scores in a Product of Experts (PoE) loss function to further improve OOD performance. With experiments conducted on natural language inference and fact verification benchmarks, we show that our method improves OOD results while maintaining in-distribution (ID) performance. 1 | Guide the Learner: Controlling Product of Experts Debiasing Method Based on Token Attribution Similarities |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.