_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d215548718 | Cross-domain sentiment analysis (CDSA) helps to address the problem of data scarcity in scenarios where labelled data for a domain (known as the target domain) is unavailable or insufficient. However, the decision to choose a domain (known as the source domain) to leverage from is, at best, intuitive. In this paper, we investigate text similarity metrics to facilitate source domain selection for CDSA. We report results on 20 domains (all possible pairs) using 11 similarity metrics. Specifically, we compare CDSA performance with these metrics for different domain-pairs to enable the selection of a suitable source domain, given a target domain. These metrics include two novel metrics for evaluating domain adaptability to help source domain selection of labelled data and utilize word and sentence-based embeddings as metrics for unlabelled data. The goal of our experiments is a recommendation chart that gives the K best source domains for CDSA for a given target domain. We show that the best K source domains returned by our similarity metrics have a precision of over 50%, for varying values of K. | Recommendation Chart of Domains for Cross-Domain Sentiment Analysis: Findings of A 20 Domain Study |
d232233599 | Typical fact verification models use retrieved written evidence to verify claims. Evidence sources, however, often change over time as more information is gathered and revised. In order to adapt, models must be sensitive to subtle differences in supporting evidence. We present VITAMINC, a benchmark infused with challenging cases that require fact verification models to discern and adjust to slight factual changes. We collect over 100,000 Wikipedia revisions that modify an underlying fact, and leverage these revisions, together with additional synthetically constructed ones, to create a total of over 400,000 claim-evidence pairs. Unlike previous resources, the examples in VITAMINC are contrastive, i.e., they contain evidence pairs that are nearly identical in language and content, with the exception that one supports a given claim while the other does not. We show that training using this design increases robustness-improving accuracy by 10% on adversarial fact verification and 6% on adversarial natural language inference (NLI). Moreover, the structure of VITAMINC leads us to define additional tasks for fact-checking resources: tagging relevant words in the evidence for verifying the claim, identifying factual revisions, and providing automatic edits via factually consistent text generation. 1 | Get Your Vitamin C! Robust Fact Verification with Contrastive Evidence |
d53301865 | Language-and-vision navigation and question answering (QA) are exciting AI tasks situated at the intersection of natural language understanding, computer vision, and robotics. Researchers from all of these fields have begun creating datasets and model architectures for these domains. It is, however, not always clear if strong performance is due to advances in multimodal reasoning or if models are learning to exploit biases and artifacts of the data. We present single modality models and explore the linguistic, visual, and structural biases of these benchmarks. We find that single modality models often outperform published baselines that accompany multimodal task datasets, suggesting a need for change in community best practices moving forward. In light of this, we recommend presenting single modality baselines alongside new multimodal models to provide a fair comparison of information gained over dataset biases when considering multimodal input. | Shifting the Baseline: Single Modality Performance on Visual Navigation & QA |
d202539612 | While neural models show remarkable accuracy on individual predictions, their internal beliefs can be inconsistent across examples. In this paper, we formalize such inconsistency as a generalization of prediction error. We propose a learning framework for constraining models using logic rules to regularize them away from inconsistency. Our framework can leverage both labeled and unlabeled examples and is directly compatible with off-the-shelf learning schemes without model redesign. We instantiate our framework on natural language inference, where experiments show that enforcing invariants stated in logic can help make the predictions of neural models both accurate and consistent. | A Logic-Driven Framework for Consistency of Neural Models |
d424534 | How can the development of ideas in a scientific field be studied over time? We apply unsupervised topic modeling to the ACL Anthology to analyze historical trends in the field of Computational Linguistics from 1978 to 2006. We induce topic clusters using Latent Dirichlet Allocation, and examine the strength of each topic over time. Our methods find trends in the field including the rise of probabilistic methods starting in 1988, a steady increase in applications, and a sharp decline of research in semantics and understanding between 1978 and 2001, possibly rising again after 2001. We also introduce a model of the diversity of ideas, topic entropy, using it to show that COLING is a more diverse conference than ACL, but that both conferences as well as EMNLP are becoming broader over time. Finally, we apply Jensen-Shannon divergence of topic distributions to show that all three conferences are converging in the topics they cover. | Studying the History of Ideas Using Topic Models |
d52156292 | In this paper, we propose a new kernel-based co-occurrence measure that can be applied to sparse linguistic expressions (e.g., sentences) with a very short learning time, as an alternative to pointwise mutual information (PMI). As well as deriving PMI from mutual information, we derive this new measure from the Hilbert-Schmidt independence criterion (HSIC); thus, we call the new measure the pointwise HSIC (PHSIC). PHSIC can be interpreted as a smoothed variant of PMI that allows various similarity metrics (e.g., sentence embeddings) to be plugged in as kernels. Moreover, PHSIC can be estimated by simple and fast (linear in the size of the data) matrix calculations regardless of whether we use linear or nonlinear kernels. Empirically, in a dialogue response selection task, PHSIC is learned thousands of times faster than an RNNbased PMI while outperforming PMI in accuracy. In addition, we also demonstrate that PH-SIC is beneficial as a criterion of a data selection task for machine translation owing to its ability to give high (low) scores to a consistent (inconsistent) pair with other pairs. | Pointwise HSIC: A Linear-Time Kernelized Co-occurrence Norm for Sparse Linguistic Expressions |
d260063187 | Social media is an extremely potent tool for influencing public opinion, particularly during important events such as elections, pandemics, and national conflicts. Emotions are a crucial aspect of this influence, but detecting them accurately in the political domain is a significant challenge due to the lack of suitable emotion labels and training datasets. In this paper, we present a generalized approach to emotion detection that can be adapted to the political domain with minimal performance sacrifice. Our approach is designed to be easily integrated into existing models without the need for additional training or fine-tuning. We demonstrate the zero-shot and few-shot performance of our model on the 2017 French presidential elections and propose efficient emotion groupings that would aid in effectively analyzing influence campaigns and agendas on social media. | Adapting Emotion Detection to Analyze Influence Campaigns on Social Media |
d248986395 | Expressing natural language descriptions of structured facts or relations -data-to-text generation (D2T) -increases the accessibility of structured knowledge repositories. Previous work(Nan et al., 2020)shows that pre-trained language models (PLMs) perform remarkably well on this task after fine-tuning on a significant amount of task-specific training data. On the other hand, while auto-regressive PLMs can generalize from a few task examples, their efficacy at D2T is largely unexplored. Furthermore, we have an incomplete understanding of the limits of PLMs on D2T. In this work, we conduct an empirical study of both finetuned and auto-regressive PLMs on the DART multi-domain D2T dataset. We consider their performance as a function of the amount of task-specific data and how the data is incorporated into the models: zero and few-shot learning, and fine-tuning of model weights. In addition, we probe the limits of PLMs by measuring performance on subsets of the evaluation data: novel predicates and abstractive test examples. To improve the performance on these subsets, we investigate two techniques: providing predicate descriptions in the context and re-ranking generated candidates by information reflected in the source. Finally, we conduct a human evaluation of model errors and show that D2T generation tasks would benefit from datasets with more careful manual curation. | What Makes Data-to-Text Generation Hard for Pretrained Language Models? |
d222090359 | Recently, neural language models (LMs) have demonstrated impressive abilities in generating high-quality discourse. While many recent papers have analyzed the syntactic aspects encoded in LMs, to date, there has been no analysis of the inter-sentential, rhetorical knowledge. In this paper, we propose a method that quantitatively evaluates the rhetorical capacities of neural LMs. We examine the capacities of neural LMs understanding the rhetoric of discourse by evaluating their abilities to encode a set of linguistic features derived from Rhetorical Structure Theory (RST). Our experiments show that BERT-based LMs outperform other Transformer LMs, revealing the richer discourse knowledge in their intermediate layer representations. In addition, GPT-2 and XLNet apparently encode less rhetorical knowledge, and we suggest an explanation drawing from linguistic philosophy. Our method presents an avenue towards quantifying the rhetorical capacities of neural LMs. | Examining the rhetorical capacities of neural language models |
d189762068 | Though great progress has been made for human-machine conversation, current dialogue system is still in its infancy: it usually converses passively and utters words more as a matter of response, rather than on its own initiatives. In this paper, we take a radical step towards building a human-like conversational agent: endowing it with the ability of proactively leading the conversation (introducing a new topic or maintaining the current topic). To facilitate the development of such conversation systems, we create a new dataset named DuConv where one acts as a conversation leader and the other acts as the follower. The leader is provided with a knowledge graph and asked to sequentially change the discussion topics, following the given conversation goal, and meanwhile keep the dialogue as natural and engaging as possible. DuConv enables a very challenging task as the model needs to both understand dialogue and plan over the given knowledge graph. We establish baseline results on this dataset (about 270K utterances and 30k dialogues) using several state-of-the-art models. Experimental results show that dialogue models that plan over the knowledge graph can make full use of related knowledge to generate more diverse multi-turn conversations. The baseline systems along with the dataset are publicly available 1 . | Proactive Human-Machine Conversation with Explicit Conversation Goals |
d243865140 | Domain adaption for word segmentation and POS tagging is a challenging problem for Chinese lexical processing. Self-training is one promising solution for it, which struggles to construct a set of high-quality pseudo training instances for the target domain. Previous work usually assumes a universal sourceto-target adaption to collect such pseudo corpus, ignoring the different gaps from the target sentences to the source domain. In this work, we start from joint word segmentation and POS tagging, presenting a fine-grained domain adaption method to model the gaps accurately. We measure the gaps by one simple and intuitive metric, and adopt it to develop a pseudo target domain corpus based on finegrained subdomains incrementally. A novel domain-mixed representation learning model is proposed accordingly to encode the multiple subdomains effectively. The whole process is performed progressively for both corpus construction and model training. Experimental results on a benchmark dataset show that our method can gain significant improvements over a vary of baselines. Extensive analyses are performed to show the advantages of our final domain adaption model as well. | A Fine-Grained Domain Adaption Model for Joint Word Segmentation and POS Tagging |
d235352728 | Few-shot relation extraction (FSRE) is of great importance in long-tail distribution problem, especially in special domain with low-resource data.Most existing FSRE algorithms fail to accurately classify the relations merely based on the information of the sentences together with the recognized entity pairs, due to limited samples and lack of knowledge. To address this problem, in this paper, we proposed a novel entity CONCEPTenhanced FEw-shot Relation Extraction scheme (ConceptFERE), which introduces the inherent concepts of entities to provide clues for relation prediction and boost the relations classification performance. Firstly, a conceptsentence attention module is developed to select the most appropriate concept from multiple concepts of each entity by calculating the semantic similarity between sentences and concepts. Secondly, a self-attention based fusion module is presented to bridge the gap of concept embedding and sentence embedding from different semantic spaces. Extensive experiments on the FSRE benchmark dataset FewRel have demonstrated the effectiveness and the superiority of the proposed ConceptFERE scheme as compared to the state-of-the-art baselines. Code is available at https://github.com/LittleGuoKe/ConceptFERE. | Entity Concept-enhanced Few-shot Relation Extraction |
d211259041 | Transformer-based models have brought a radical change to neural machine translation. A key feature of the Transformer architecture is the so-called multi-head attention mechanism, which allows the model to focus simultaneously on different parts of the input. However, recent works have shown that most attention heads learn simple, and often redundant, positional patterns. In this paper, we propose to replace all but one attention head of each encoder layer with simple fixed -non-learnable -attentive patterns that are solely based on position and do not require any external knowledge. Our experiments with different data sizes and multiple language pairs show that fixing the attention heads on the encoder side of the Transformer at training time does not impact the translation quality and even increases BLEU scores by up to 3 points in low-resource scenarios. | Fixed Encoder Self-Attention Patterns in Transformer-Based Machine Translation |
d256461140 | This paper describes the systems submitted by BFCAI team to Nuanced Arabic Dialect Identification (NADI) shared task 2022. Dialect identification task aims at detecting the source variant of a given text or speech segment automatically. There are two subtasks in NADI 2022, the first subtask for country-level identification and the second subtask for sentiment analysis. Our team participated in the first subtask. The proposed systems use Term Frequency Inverse/Document Frequency and word embeddings as vectorization models. Different machine learning algorithms have been used as classifiers. The proposed systems have been tested on two test sets: Test-A and Test-B. The proposed models achieved Macro-f1 score of 21.25% and 9.71% for Test-A and Test-B set respectively. On other hand, the best-performed submitted system achieved Macro-f1 score of 36.48% and 18.95% for Test-A and Test-B set respectively. | Word Representation Models for Arabic Dialect Identification |
d259833796 | This paper presents our system for the MEDIQA-Chat 2023 shared task on medical conversation summarization. Our approach involves finetuning a LongT5 model on multiple tasks simultaneously, which we demonstrate improves the model's overall performance while reducing the number of factual errors and hallucinations in the generated summary. Furthermore, we investigated the effect of augmenting the data with in-text annotations from a clinical named entity recognition model, finding that this approach decreased summarization quality. Lastly, we explore using different text generation strategies for medical note generation based on the length of the note. Our findings suggest that the application of our proposed approach can be beneficial for improving the accuracy and effectiveness of medical conversation summarization. | Calvados at MEDIQA-Chat 2023: Improving Clinical Note Generation with Multi-Task Instruction Finetuning |
d259833860 | We propose a method to automate orthodontic diagnosis with natural language processing. It is worthwhile to assist dentists with such technology to prevent errors by inexperienced dentists and to reduce the workload of experienced ones. However, text length and style inconsistencies in medical findings make an automated orthodontic diagnosis with deep-learning models difficult. In this study, we improve the performance of automatic diagnosis utilizing short summaries of medical findings written in a consistent style by experienced dentists. Experimental results on 970 Japanese medical findings show that summarization consistently improves the performance of various machine learning models for automated orthodontic diagnosis. Although BERT is the model that gains the most performance with the proposed method, the convolutional neural network achieved the best performance. | Automated Orthodontic Diagnosis from a Summary of Medical Findings |
d259370834 | Parameter-Efficient Tuning (PET) has shown remarkable performance by fine-tuning only a small number of parameters of the pre-trained language models (PLMs) for the downstream tasks, while it is also possible to construct backdoor attacks due to the vulnerability of pretrained weights. However, a large reduction in the number of attackable parameters in PET will cause the user's fine-tuning to greatly affect the effectiveness of backdoor attacks, resulting in backdoor forgetting. We find that the backdoor injection process can be regarded as multitask learning, which has a convergence imbalance problem between the training of clean and poisoned data. And this problem might result in forgetting the backdoor. Based on this finding, we propose a gradient control method to consolidate the attack effect, comprising two strategies. One controls the gradient magnitude distribution cross layers within one task and the other prevents the conflict of gradient directions between tasks. Compared with previous backdoor attack methods in the scenario of PET, our method improves the effect of the attack on sentiment classification and spam detection respectively, which shows that our method is widely applicable to different tasks. | A Gradient Control Method for Backdoor Attacks on Parameter-Efficient Tuning |
d222290439 | Fact verification models have enjoyed a fast advancement in the last two years with the development of pre-trained language models like BERT and the release of large scale datasets such as FEVER. However, the challenging problem of fake news detection has not benefited from the improvement of fact verification models, which is closely related to fake news detection. In this paper, we propose a simple yet effective approach to connect the dots between fact verification and fake news detection. Our approach first employs a text summarization model pre-trained on news corpora to summarize the long news article into a short claim. Then we use a fact verification model pre-trained on the FEVER dataset to detect whether the input news article is real or fake. Our approach makes use of the recent success of fact verification models and enables zero-shot fake news detection, alleviating the need of large scale training data to train fake news detection models. Experimental results on FakenewsNet, a benchmark dataset for fake news detection, demonstrate the effectiveness of our proposed approach. * Equal Contribution. This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http:// creativecommons.org/licenses/by/4.0/. | Connecting the Dots Between Fact Verification and Fake News Detection |
d207847777 | Cross-lingual word embeddings transfer knowledge between languages: models trained on high-resource languages can predict in low-resource languages. We introduce CLIME, an interactive system to quickly refine cross-lingual word embeddings for a given classification problem. First, CLIME ranks words by their salience to the downstream task. Then, users mark similarity between keywords and their nearest neighbors in the embedding space. Finally, CLIME updates the embeddings using the annotations. We evaluate CLIME on identifying health-related text in four low-resource languages: Ilocano, Sinhalese, Tigrinya, and Uyghur. Embeddings refined by CLIME capture more nuanced word semantics and have higher test accuracy than the original embeddings. CLIME often improves accuracy faster than an active learning baseline and can be easily combined with active learning to improve results. | Interactive Refinement of Cross-Lingual Word Embeddings |
d11910890 | Using automatic measures such as labeled and unlabeled attachment scores is common practice in dependency parser evaluation. In this paper, we examine whether these measures correlate with human judgments of overall parse quality. We ask linguists with experience in dependency annotation to judge system outputs. We measure the correlation between their judgments and a range of parse evaluation metrics across five languages. The humanmetric correlation is lower for dependency parsing than for other NLP tasks. Also, inter-annotator agreement is sometimes higher than the agreement between judgments and metrics, indicating that the standard metrics fail to capture certain aspects of parse quality, such as the relevance of root attachment or the relative importance of the different parts of speech. | Do dependency parsing metrics correlate with human judgments? |
d32260611 | What would be a good method to provide a large collection of semantically annotated texts with formal, deep semantics rather than shallow? In this talk I will argue that (i) a bootstrapping approach comprising state-of-the-art NLP tools for semantic parsing, in combination with (ii) a wiki-like interface for collaborative annotation of experts, and (iii) a game with a purpose for crowdsourcing, are the starting ingredients for fulfilling this enterprise. The result, known as the Groningen Meaning Bank, is a semantic resource that anyone can edit and that integrates various semantic phenomena, including predicate-argument structure, scope, tense, thematic roles, animacy, pronouns, and rhetorical relations. A single semantic formalism, Discourse Representation Theory, embraces all these phenonema by taking meaning representations of texts rather than sentences as the units of annotation. | The Groningen Meaning Bank |
d218596311 | Multiword expressions (MWEs) represent lexemes that should be treated as single lexical units due to their idiosyncratic nature. Multiple NLP applications have been shown to benefit from MWE identification, however the research on lexical complexity of MWEs is still an under-explored area. In this work, we re-annotate the Complex Word Identification Shared Task 2018 dataset of Yimam et al.(2017), which provides complexity scores for a range of lexemes, with the types of MWEs. We release the MWE-annotated dataset with this paper, and we believe this dataset represents a valuable resource for the text simplification community. In addition, we investigate which types of expressions are most problematic for native and non-native readers. Finally, we show that a lexical complexity assessment system benefits from the information about MWE types. | Detecting Multiword Expression Type Helps Lexical Complexity Assessment |
d6193081 | In this paper, we propose a multi-step stacked learning model for disfluency detection. Our method incorporates refined n-gram features step by step from different word sequences. First, we detect filler words. Second, edited words are detected using n-gram features extracted from both the original text and filler filtered text. In the third step, additional n-gram features are extracted from edit removed texts together with our newly induced in-between features to improve edited word detection. We use Max-Margin Markov Networks (M 3 Ns) as the classifier with the weighted hamming loss to balance precision and recall. Experiments on the Switchboard corpus show that the refined n-gram features from multiple steps and M 3 Ns with weighted hamming loss can significantly improve the performance. Our method for disfluency detection achieves the best reported F-score 0.841 without the use of additional resources. 1 | Disfluency Detection Using Multi-step Stacked Learning |
d9909120 | The traditional method to evaluate a knowledge extraction system is to measure precision and recall. But this method only partially measures the quality of a knowledge base (KB) as it cannot predict whether a KB is useful or not. One of the ways in which a KB can be useful is if it is able to deduce implicit information from text which standard information extraction techniques cannot extract. We propose a novel, simple evaluation framework called "Machine Reading between the Lines" (MRbtL) which measures the usefulness of extracted KBs by determining how much they can help improve a relation extraction system. In our experiments, we compare two KBs which both have high precision and recall according to annotators who evaluate the knowledge in the KBs independently from any application or context. But, we show that one outperforms the other in terms of MRbtL experiments, as it can accurately deduce more new facts from the output of a relation extractor more accurately. In short, one extracted KB can read between the lines to identify extra information, whereas the other one cannot. | Machine Reading Between the Lines: A Simple Evaluation Framework for Extracted Knowledge Bases |
d274239 | Humor identification is a hard natural language understanding problem. We identify a subproblem -the "that's what she said" problem -with two distinguishing characteristics: (1) use of nouns that are euphemisms for sexually explicit nouns and (2) structure common in the erotic domain. We address this problem in a classification approach that includes features that model those two characteristics. Experiments on web data demonstrate that our approach improves precision by 12% over baseline techniques that use only word-based features. | That's What She Said: Double Entendre Identification |
d199552196 | We introduce EASSE, a Python package aiming to facilitate and standardise automatic evaluation and comparison of Sentence Simplification (SS) systems. EASSE provides a single access point to a broad range of evaluation resources: standard automatic metrics for assessing SS outputs (e.g. SARI), wordlevel accuracy scores for certain simplification transformations, reference-independent quality estimation features (e.g. compression ratio), and standard test data for SS evaluation (e.g. TurkCorpus). Finally, EASSE generates easy-to-visualise reports on the various metrics and features above and on how a particular SS output fares against reference simplifications. Through experiments, we show that these functionalities allow for better comparison and understanding of the performance of SS systems. | EASSE: Easier Automatic Sentence Simplification Evaluation |
d258486872 | We present a graph-based tool which can be used to explore Verbal Multi-Word Expression (VMWE) annotated in the Parseme project. The tool can be used for linguistic exploration on the data, for helping the manual annotation process and to search for errors or inconsistencies in the annotations. | Graph-based multi-layer querying in Parseme Corpora |
d14547126 | This paper describes SimpleNLG, a realisation engine for English which aims to provide simple and robust interfaces to generate syntactic structures and linearise them. The library is also flexible in allowing the use of mixed (canned and noncanned) representations. | SimpleNLG: A realisation engine for practical applications |
d222341640 | Extrapolation to unseen sequence lengths is a challenge for neural generative models of language. In this work, we characterize the effect on length extrapolation of a modeling decision often overlooked: predicting the end of the generative process through the use of a special end-of-sequence (EOS) vocabulary item. We study an oracle setting-forcing models to generate to the correct sequence length at test time-to compare the lengthextrapolative behavior of networks trained to predict EOS (+EOS) with networks not trained to (-EOS). We find that -EOS substantially outperforms +EOS, for example extrapolating well to lengths 10 times longer than those seen at training time in a bracket closing task, as well as achieving a 40% improvement over +EOS in the difficult SCAN dataset length generalization task. By comparing the hidden states and dynamics of -EOS and +EOS models, we observe that +EOS models fail to generalize because they (1) unnecessarily stratify their hidden states by their linear position is a sequence (structures we call length manifolds) or (2) get stuck in clusters (which we refer to as length attractors) once the EOS token is the highest-probability prediction.Gary Marcus. 2018. Deep learning: A critical appraisal. arXiv preprint arXiv:1801.00631.William Merrill. 2020. On the linguistic capacity of real-time counter automata. arXiv preprint arXiv:2004.06866. | The EOS Decision and Length Extrapolation |
d131763622 | Popular Natural Language Inference (NLI) datasets have been shown to be tainted by hypothesis-only biases. Adversarial learning may help models ignore sensitive biases and spurious correlations in data. We evaluate whether adversarial learning can be used in NLI to encourage models to learn representations free of hypothesis-only biases. Our analyses indicate that the representations learned via adversarial learning may be less biased, with only small drops in NLI accuracy. | On Adversarial Removal of Hypothesis-only Bias in Natural Language Inference |
d219573466 | As conversational AI-based dialogue management has increasingly become a trending topic, the need for a standardized and reliable evaluation procedure grows even more pressing. The current state of affairs suggests various evaluation protocols to assess chat-oriented dialogue management systems, rendering it difficult to conduct fair comparative studies across different approaches and gain an insightful understanding of their values. To foster this research, a more robust evaluation protocol must be set in place. This paper presents a comprehensive synthesis of both automated and human evaluation methods on dialogue systems, identifying their shortcomings while accumulating evidence towards the most effective evaluation dimensions. A total of 20 papers from the last two years are surveyed to analyze three types of evaluation protocols: automated, static, and interactive. Finally, the evaluation dimensions used in these papers are compared against our expert evaluation on the system-user dialogue data collected from the Alexa Prize 2020. | Towards Unified Dialogue System Evaluation: A Comprehensive Analysis of Current Evaluation Protocols |
d256461358 | Social media platforms are becoming inherent parts of people's daily life to express opinions and stances toward topics of varying polarities. Stance detection determines the viewpoint expressed in a text toward a target. While communication on social media (e.g., Twitter) takes place in more than 40 languages, the majority of stance detection research has been focused on English. Although some efforts have recently been made to develop stance detection datasets in other languages, no similar efforts seem to have considered the Arabic language. In this paper, we present MAWQIF, the first Arabic dataset for target-specific stance detection, composed of 4,121 tweets annotated with stance, sentiment, and sarcasm polarities. MAWQIF, as a multi-label dataset, can provide more opportunities for studying the interaction between different opinion dimensions and evaluating a multi-task model. We provide a detailed description of the dataset, present an analysis of the produced annotation, and evaluate four BERT-based models on it. Our best model achieves a macro-F 1 of 78.89%, which shows that there is ample room for improvement on this challenging task. We publicly release our dataset, the annotation guidelines, and the code of the experiments. 1 | MAWQIF: A Multi-label Arabic Dataset for Target-specific Stance Detection |
d258378243 | Large neural language models (LLMs) can be powerful tools for research in lexical semantics. We illustrate this potential using the English verb break, which has numerous senses and appears in a wide range of syntactic frames. We show that LLMs capture known sense distinctions and can be used to identify informative new sense combinations for further analysis. More generally, we argue that LLMs are aligned with lexical semantic theories in providing high-dimensional, contextually modulated representations, but LLMs' lack of discrete features and dependence on usage-based data offer a genuinely new perspective on traditional problems in lexical semantics. * Data and code available at https://github.com/ epetsen/break-llms. | Lexical Semantics with Large Language Models: A Case Study of English break * |
d258378250 | Pre-trained models usually come with a predefined tokenization and little flexibility as to what subword tokens can be used in downstream tasks. This problem concerns especially multilingual NLP and low-resource languages, which are typically processed using cross-lingual transfer. In this paper, we aim to find out if the right granularity of tokenization is helpful for a text classification task, namely dialect classification. Aiming at generalizations beyond the studied cases, we look for the optimal granularity in four dialect datasets, two with relatively consistent writing (one Arabic and one Indo-Aryan set) and two with considerably inconsistent writing (one Arabic and one Swiss German set). To gain more control over subword tokenization and ensure direct comparability in the experimental settings, we train a CNN classifier from scratch comparing two subword tokenization methods (Unigram model and BPE). For reference, we compare the results obtained in our analysis to the state of the art achieved by fine-tuning pre-trained models. We show that models trained from scratch with an optimal tokenization level perform better than fine-tuned classifiers in the case of highly inconsistent writing. In the case of relatively consistent writing, fine-tuned models remain better regardless of the tokenization level. 1 | Optimizing the Size of Subword Vocabularies in Dialect Classification |
d257985704 | Clinical concept normalization involves linking entity mentions in clinical narratives to their corresponding concepts in standardized medical terminologies. It can be used to determine the specific meaning of a mention, facilitating effective use and exchange of clinical information, and to support semantic cross-compatibility of texts. We present a rule-based multipass sieve approach incorporating both exact and approximate matching based on dictionaries, and experiment with back-translation as a means of data augmentation. The dictionaries are built from the UMLS Metathesaurus as well as MCN corpus training data. Additionally, we train a multi-class baseline based on BERT. Our multi-pass sieve approach achieves an accuracy of 82.0% on the MCN corpus, the highest for any rule-based method. A hybrid method combining these two achieves a slightly higher accuracy of 82.3%.RÉSUMÉ. La normalisation des concepts cliniques consiste à relier les mentions d'entités dans les récits cliniques à leurs concepts correspondants dans des terminologies médicales normalisées. Il peut étre utilisé pour déterminer la signification spécifique d'une mention, faciliter l'utilisation et l'échange efficaces d'informations cliniques et soutenir la compatibilité sémantique des textes. Nous présentons une approche de tamisage multi-passes intégrant deux types de correspondance -exacte et approximative -basée sur des dictionnaires construits avec UMLS Metathesaurus et le corpus MCN, et expérimentons la rétro-traduction comme moyen d'augmenter les données. De plus, nous preparons une méthode de référence multi-classes basée sur BERT. Notre méthode de tamisage multi-passes atteint une précision de 82,0% sur le corpus MCN, la plus élevée de toutes les méthodes fondée sur des règles. Notre méthode hybride réalise une précision légèrement supérieure de 82,3%. | A Multi-pass Sieve for Clinical Concept Normalization |
d233189560 | We introduce a general framework for abstractive summarization with factual consistency and distinct modeling of the narrative flow in an output summary. Our work addresses current limitations of models for abstractive summarization that often hallucinate information or generate summaries with coherence issues.To generate abstractive summaries with factual consistency and narrative flow, we propose Cooperative Generator -Discriminator Networks (Co-opNet), a novel transformerbased framework where a generator works with a discriminator architecture to compose coherent long-form summaries. We explore four different discriminator objectives which each capture a different aspect of coherence, including whether salient spans of generated abstracts are hallucinated or appear in the input context, and the likelihood of sentence adjacency in generated abstracts.We measure the ability of Co-opNet to learn these objectives with arXiv scientific papers, using the abstracts as a proxy for gold longform scientific article summaries. Empirical results from automatic and human evaluations demonstrate that Co-opNet learns to summarize with considerably improved global coherence compared to competitive baselines. | Discourse Understanding and Factual Consistency in Abstractive Summarization |
d210861173 | This paper describes "TLT-school", a corpus of speech utterances collected in schools of northern Italy for assessing the performance of students learning both English and German. The corpus was recorded in the years 2017 and 2018 from students aged between nine and sixteen years, attending primary, middle and high school. All utterances have been scored, in terms of some predefined proficiency indicators, by human experts. In addition, most of utterances recorded in 2017 have been manually transcribed carefully. Guidelines and procedures used for manual transcriptions of utterances will be described in detail, as well as results achieved by means of an automatic speech recognition system developed by us. Part of the corpus is going to be freely distributed to scientific community particularly interested both in non-native speech recognition and automatic assessment of second language proficiency. | TLT-school: a Corpus of Non Native Children Speech |
d236486208 | Expressions with an aspectual variant of a light verb, e.g. take on debt vs. have debt, are frequent in texts but often difficult to classify between verbal idioms, light verb constructions or compositional phrases. We investigate the properties of such expressions with a disputed membership and propose a selection of features that determine more satisfactory boundaries between the three categories in this zone, assigning the expressions to one of them. | Where Do Aspectual Variants of Light Verb Constructions Belong? |
d222310232 | To assist human review process, we build a novel ReviewRobot to automatically assign a review score and write comments for multiple categories such as novelty and meaningful comparison. A good review needs to be knowledgeable, namely that the comments should be constructive and informative to help improve the paper; and explainable by providing detailed evidence. ReviewRobot achieves these goals via three steps: (1) We perform domainspecific Information Extraction to construct a knowledge graph (KG) from the target paper under review, a related work KG from the papers cited by the target paper, and a background KG from a large collection of previous papers in the domain. (2) By comparing these three KGs, we predict a review score and detailed structured knowledge as evidence for each review category. (3) We carefully select and generalize human review sentences into templates, and apply these templates to transform the review scores and evidence into natural language comments. Experimental results show that our review score predictor reaches 71.4%-100% accuracy. Human assessment by domain experts shows that 41.7%-70.5% of the comments generated by ReviewRobot are valid and constructive, and better than humanwritten ones for 20% of the time. Thus, Re-viewRobot can serve as an assistant for paper reviewers, program chairs and authors. 1 | ReviewRobot: Explainable Paper Review Generation based on Knowledge Synthesis |
d237431351 | We aim to automatically identify human action reasons in online videos. We focus on the widespread genre of lifestyle vlogs, in which people perform actions while verbally describing them. We introduce and make publicly available the WHYACT dataset, consisting of 1,077 visual actions manually annotated with their reasons. We describe a multimodal model that leverages visual and textual information to automatically infer the reasons corresponding to an action presented in the video. | WHYACT: Identifying Action Reasons in Lifestyle Vlogs |
d216080844 | Deep reinforcement learning is a promising approach to training a dialog manager, but current methods struggle with the large state and action spaces of multi-domain dialog systems. Building upon Deep Q-learning from Demonstrations (DQfD), an algorithm that scores highly in difficult Atari games, we leverage dialog data to guide the agent to successfully respond to a user's requests. We make progressively fewer assumptions about the data needed, using labeled, reduced-labeled, and even unlabeled data to train expert demonstrators. We introduce Reinforced Fine-tune Learning, an extension to DQfD, enabling us to overcome the domain gap between the datasets and the environment. Experiments in a challenging multi-domain dialog system framework validate our approaches, and get high success rates even when trained on outof-domain data. | Learning Dialog Policies from Weak Demonstrations |
d259833851 | Modern machine learning techniques have produced many impressive results in language technology, but these techniques generally require an amount of training data that is many orders of magnitude greater than what exists for low-resource languages in general, and endangered languages in particular. However, dictionary definitions in a comparatively much more well-resourced majority language can provide a link between low-resource languages and machine learning models trained on massive amounts of majority-language training data. Promising results have been achieved by leveraging these embeddings in the search mechanisms of bilingual dictionaries of Plains Cree (nêhiyawêwin), Arapaho (Hinóno'éitíit), Northern Haida (Xaad Kíl), and Tsuut'ina (Tsúùt'ínà), four Indigenous languages spoken in North America. Not only are the search results in the majority language of the definitions more relevant, but they can be semantically relevant in ways not achievable with classic information retrieval techniques: users can perform successful searches for words that do not occur at all in the dictionary. Not only this, but these techniques are directly applicable to any bilingual dictionary providing translations between a high-and low-resource language. | Finding words that aren't there: Using word embeddings to improve dictionary search for low-resource languages |
d9420249 | Recently, categorical grammar has been focused as a powerful grammar. This paper aims to develop a framework for automatic CG tagging for Thai. We investigated two main algorithms, CRF and Statistical alignment model based on information theory (SAM). We found that SAM gives the best results both in word level and sentence level. We got the accuracy 89.25% in word level and 82.49% in sentence level. SAM is better than CRF in known word. On the other hand, CRF is better than SAM when we applied for unknown word. Combining both methods can be suited for both known and unknown word. | AutoTagTCG : A Framework for Automatic Thai CG Tagging |
d224704623 | Much as the social landscape in which languages are spoken shifts, language too evolves to suit the needs of its users. Lexical semantic change analysis is a burgeoning field of semantic analysis which aims to trace changes in the meanings of words over time. This paper presents an approach to lexical semantic change detection based on Bayesian word sense induction suitable for novel word sense identification. This approach is used for a submission to SemEval-2020 Task 1, which shows the approach to be capable of the SemEval task. The same approach is also applied to a corpus gleaned from 15 years of Twitter data, the results of which are then used to identify words which may be instances of slang.Related WorkApproaches to lexical semantic change detection collect co-occurrence information on words, represented in most approaches by pointwise mutual information scores(Tahmasebi et al., 2018). This information may be used as the basis of an approach, as is used bySagi et al. (2009)and Gulordava and BaroniAccepted for publication at SemEval 2020 This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: | UoB at SemEval-2020 Task 1: Automatic Identification of Novel Word Senses |
d213004368 | Named entity recognition is a natural language processing task to recognize and extract spans of text associated with named entities and classify them in semantic Categories.Google BERT is a deep bidirectional language model, pre-trained on large corpora that can be fine-tuned to solve many NLP tasks such as question answering, named entity recognition, part of speech tagging and etc. In this paper, we use the pre-trained deep bidirectional network, BERT, to make a model for named entity recognition in Persian.We also compare the results of our model with the previous state of the art results achieved on Persian NER. Our evaluation metric is CONLL 2003 score in two levels of word and phrase. This model achieved second place in NSURL-2019 task 7 competition which associated with NER for the Persian language. our results in this competition are 83.5 and 88.4 f1 CONLL score respectively in phrase and word level evaluation. | Beheshti-NER: Persian named entity recognition Using BERT |
d928331 | We have aligned Japanese and English news articles and sentences to make a large parallel corpus. We first used a method based on cross-language information retrieval (CLIR) to align the Japanese and English articles and then used a method based on dynamic programming (DP) matching to align the Japanese and English sentences in these articles. However, the results included many incorrect alignments. To remove these, we propose two measures (scores) that evaluate the validity of alignments. The measure for article alignment uses similarities in sentences aligned by DP matching and that for sentence alignment uses similarities in articles aligned by CLIR. They enhance each other to improve the accuracy of alignment. Using these measures, we have successfully constructed a largescale article and sentence alignment corpus available to the public. | Reliable Measures for Aligning Japanese-English News Articles and Sentences |
d245838348 | ||
d198909640 | Both statistical (SMT) and neural (NMT) approaches to machine translation (MT) explore large search spaces to produce and score translations. It is however well known that often the top hypothesis as scored by such approaches may not be the best overall translation among those that can be produced. Previous work on SMT has extensively explored re-ranking strategies in attempts to find the best possible translation. In this paper, we focus on NMT and provide an in-depth investigation to explore the influence of beam sizes on information content and translation quality. We gather new insights using oracle experiments on the efficacy of exploiting larger beams and propose a simple, yet novel consensus-based, n-best re-ranking approach that makes use of different automatic evaluation metrics to measure consensus in n-best lists. Our results reveal that NMT is able to cover more of the information content of the references compared to SMT and that this leads to better re-ranked translations (according to human evaluation). We further show that the MT evaluation metric used for the consensus-based re-ranking plays a major role, with character-based metrics performing better than BLEU. | Exploring Hypotheses Spaces in Neural Machine Translation |
d237347097 | Content Warning: This paper contains examples of stereotypes and associations, misgendering, erasure, and other harms that could be offensive and triggering to trans and nonbinary individuals.Gender is widely discussed in the context of language tasks and when examining the stereotypes propagated by language models. However, current discussions primarily treat gender as binary, which can perpetuate harms such as the cyclical erasure of non-binary gender identities. These harms are driven by model and dataset biases, which are consequences of the non-recognition and lack of understanding of non-binary genders in society. In this paper, we explain the complexity of gender and language around it, and survey non-binary persons to understand harms associated with the treatment of gender as binary in English language technologies. We also detail how current language representations (e.g., GloVe, BERT) capture and perpetuate these harms and related challenges that need to be acknowledged and addressed for representations to equitably encode gender information. | Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologies Sunipa Dev she/her UCLA Masoud Monajatipoor* he/him UCLA Anaelia Ovalle* they/he/she UCLA Arjun Subramonian* they/them UCLA, Queer in AI |
d160019345 | In many machine learning scenarios, supervision by gold labels is not available and consequently neural models cannot be trained directly by maximum likelihood estimation. In a weak supervision scenario, metric-augmented objectives can be employed to assign feedback to model outputs, which can be used to extract a supervision signal for training. We present several objectives for two separate weakly supervised tasks, machine translation and semantic parsing. We show that objectives should actively discourage negative outputs in addition to promoting a surrogate gold structure. This notion of bipolarity is naturally present in ramp loss objectives, which we adapt to neural models. We show that bipolar ramp loss objectives outperform other non-bipolar ramp loss objectives and minimum risk training on both weakly supervised tasks, as well as on a supervised machine translation task. Additionally, we introduce a novel token-level ramp loss objective, which is able to outperform even the best sequence-level ramp loss on both weakly supervised tasks.233 | Learning Neural Sequence-to-Sequence Models from Weak Feedback with Bipolar Ramp Loss |
d227230841 | Globalization has caused the rise of the code-switching phenomenon among multilingual societies. In Arab countries, code-switching between Arabic and English has become frequent, especially through social media platforms. Consequently, research in Natural Language Processing (NLP) systems increased to tackle such a phenomenon. One of the significant challenges of developing code-switched NLP systems is the lack of data itself. In this paper, we propose an open source trained bilingual contextual word embedding models of FLAIR, BERT, and ELEC-TRA. We also propose a novel contextual word embedding model called KERMIT, which can efficiently map Arabic and English words inside one vector space in terms of data usage. We applied intrinsic and extrinsic evaluation methods to compare the performance of the models. Our results show that FLAIR and FastText achieve the highest results in the sentiment analysis task. However, KERMIT is the best-achieving model on the intrinsic evaluation and named entity recognition. Also, it outperforms the other transformer-based models on question answering task.This work is licensed under a Creative Commons Attribution 4.0 International License. License details: http://creativecommons.org/licenses/by/4.0/. | Contextual Embeddings for Arabic-English Code-Switched Data |
d12418716 | The goal of this paper is to describe the annotation protocols and the Semantic Annotation Tool (SAT) used in the DutchSemCor project. The DutchSemCor project is aiming at aligning the Cornetto lexical database with the Dutch language corpus SoNaR. 250K corpus occurrences of the 3,000 most frequent and most ambiguous Dutch nouns, adjectives and verbs are being annotated manually using the SAT. This data is then used for bootstrapping 750K extra occurrences which in turn will be checked manually. Our main focus in this paper is the methodology applied in the project to attain the envisaged Inter-annotator Agreement (IA) of 80%. We will also discuss one of the main objectives of DutchSemCor i.e. to provide semantically annotated language data with high scores for quantity, quality and diversity. Sample data with high scores for these three features can yield better results for co-training WSD systems. Finally, we will take a brief look at our annotation tool. | Computer assisted semantic annotation in the DutchSemCor project |
d252367868 | The task of completing knowledge triplets has broad downstream applications. Both structural and semantic information plays an important role in knowledge graph completion. Unlike previous approaches that rely on either the structures or semantics of the knowledge graphs, we propose to jointly embed the semantics in the natural language description of the knowledge triplets with their structure information. Our method embeds knowledge graphs for the completion task via fine-tuning pre-trained language models with respect to a probabilistic structured loss, where the forward pass of the language models captures semantics and the loss reconstructs structures. Our extensive experiments on a variety of knowledge graph benchmarks have demonstrated the state-of-the-art performance of our method. We also show that our method can significantly improve the performance in a low-resource regime, thanks to the better use of semantics. The code and datasets are available at https://github.com/pkusjh/LASS. | Joint Language Semantic and Structure Embedding for Knowledge Graph Completion |
d222134166 | Traditional (unstructured) pruning methods for a Transformer model focus on regularizing the individual weights by penalizing them toward zero. In this work, we explore spectralnormalized identity priors (SNIP), a structured pruning approach that penalizes an entire residual module in a Transformer model toward an identity mapping. Our method identifies and discards unimportant non-linear mappings in the residual connections by applying a thresholding operator on the function norm. It is applicable to any structured module, including a single attention head, an entire attention block, or a feed-forward subnetwork. Furthermore, we introduce spectral normalization to stabilize the distribution of the post-activation values of the Transformer layers, further improving the pruning effectiveness of the proposed methodology. We conduct experiments with BERT on 5 GLUE benchmark tasks to demonstrate that SNIP achieves effective pruning results while maintaining comparable performance. Specifically, we improve the performance over the state-of-the-art by 0.5 to 1.0% on average at 50% compression ratio. | Pruning Redundant Mappings in Transformer Models via Spectral-Normalized Identity Prior |
d2351337 | Vector-based models of lexical semantics retrieve semantically related words automatically from large corpora by exploiting the property that words with a similar meaning tend to occur in similar contexts. Despite their increasing popularity, it is unclear which kind of semantic similarity they actually capture and for which kind of words. In this paper, we use three vector-based models to retrieve semantically related words for a set of Dutch nouns and we analyse whether three linguistic properties of the nouns influence the results. In particular, we compare results from a dependency-based model with those from a 1st and 2nd order bag-of-words model and we examine the effect of the nouns' frequency, semantic speficity and semantic class. We find that all three models find more synonyms for high-frequency nouns and those belonging to abstract semantic classses. Semantic specificty does not have a clear influence. | Modelling Word Similarity. An Evaluation of Automatic Synonymy Extraction Algorithms |
d220445369 | A lot of progress has been made to improve question answering (QA) in recent years, but the special problem of QA over narrative book stories has not been explored in-depth. We formulate BookQA as an open-domain QA task given its similar dependency on evidence retrieval. We further investigate how state-ofthe-art open-domain QA approaches can help BookQA. Besides achieving state-of-the-art on the NarrativeQA benchmark, our study also reveals the difficulty of evidence retrieval in books with a wealth of experiments and analysis -which necessitates future effort on novel solutions for evidence retrieval in BookQA. | Frustratingly Hard Evidence Retrieval for QA Over Books |
d216056409 | Hindi grapheme-to-phoneme (G2P) conversion is mostly trivial, with one exception: whether a schwa represented in the orthography is pronounced or unpronounced (deleted). Previous work has attempted to predict schwa deletion in a rule-based fashion using prosodic or phonetic analysis. We present the first statistical schwa deletion classifier for Hindi, which relies solely on the orthography as the input and outperforms previous approaches. We trained our model on a newly-compiled pronunciation lexicon extracted from various online dictionaries. Our best Hindi model achieves state of the art performance, and also achieves good performance on a closely related language, Punjabi, without modification. | Supervised Grapheme-to-Phoneme Conversion of Orthographic Schwas in Hindi and Punjabi |
d11993227 | This work investigates the potential use of post-edited machine translation (MT) outputs as reference translations for automatic machine translation evaluation, focusing mainly on the following important question: Is it necessary to take into account the machine translation system and the source language from which the given post-edits are generated? In order to explore this, we investigated the use of post-edits originating from different machine translation systems (two statistical systems and two rule-based systems), as well as the use of post-edits originating from two different source languages (English and German). The obtained results shown that for comparison of different systems using automatic evaluation metrics, a good option is to use a post-edit originating from a high-quality (possibly distinct) system. A better option is to use it together with other references and post-edits, however post-edits originating from poor translation systems should be avoided. For tuning or development of a particular system, post-edited output of this same system seems to be the best reference translation. | Title Potential and limits of using post-edits as reference translations for MT evaluation Potential and Limits of Using Post-edits as Reference Translations for MT Evaluation |
d247434179 | Brain encoding aims at reconstructing fMRI brain activity given a stimulus. Earlier neural encoding models focused on brain encoding for single-mode stimuli: visual (pretrained CNNs) or text (pretrained language models). Few recent papers have also obtained separate visual and text representation models and performed late-fusion using simple heuristics. However, the human brain perceives the environment using information from multiple modalities, and previous works have not explored the co-attentive multi-modal encoding for visual and text reasoning. This paper systematically explores image and multi-modal Transformers' efficacy for brain encoding. Extensive experiments on two popular datasets, BOLD5000 and Pereira, provide the following insights. (1) We find that VisualBERT, a multi-modal Transformer, significantly outperforms previously proposed single-mode CNNs, image Transformers, and other previously proposed multi-modal models, thereby establishing new state-of-the-art. (2) The regions such as LPTG, LMTG, LIFG, and STS, which have dual functionalities for language and vision, have a higher correlation with multi-modal models, which reinforces the fact that these models are good at mimicking the human brain behavior. (3) The supremacy of visio-linguistic models raises the question of whether the responses elicited in the visual regions are affected implicitly by linguistic processing even when passively viewing images. Future fMRI tasks can verify this computational insight in an appropriate experimental setting. We make our code publicly available 1 . | Visio-Linguistic Brain Encoding |
d235421836 | Knowledge Graph (KG) completion research usually focuses on densely connected benchmark datasets that are not representative of real KGs. We curate two KG datasets that include biomedical and encyclopedic knowledge and use an existing commonsense KG dataset to explore KG completion in the more realistic setting where dense connectivity is not guaranteed. We develop a deep convolutional network that utilizes textual entity representations and demonstrate that our model outperforms recent KG completion methods in this challenging setting. We find that our model's performance improvements stem primarily from its robustness to sparsity. We then distill the knowledge from the convolutional network into a student network that re-ranks promising candidate entities. This re-ranking stage leads to further improvements in performance and demonstrates the effectiveness of entity re-ranking for KG completion. 1 | Robust Knowledge Graph Completion with Stacked Convolutions and a Student Re-Ranking Network |
d261341898 | ||
d38166371 | In this paper we present a linguistic resource for the lexical representation of affective knowledge. This resource (named WORDNET-AFFECT) was developed starting from WORDNET, through a selection and tagging of a subset of synsets representing the affective meanings. | WordNet-Affect: an Affective Extension of WordNet |
d202766228 | We introduce Tanbih, a news aggregator with intelligent analysis tools to help readers understanding what is behind a news story. Our system displays news grouped into events and generates media profiles that show the general factuality of reporting, the degree of propagandistic content, hyper-partisanship, leading political ideology, general frame of reporting, and stance with respect to various claims and topics of a news outlet. In addition, we automatically analyze each article in order to detect whether it is propagandistic and to determine its stance with respect to a number of controversial topics. | Tanbih: Get To Know What You Are Reading |
d253761978 | The current classification methods for relation extraction (RE) generally utilize pre-trained language models (PLMs) and have achieved superior results. However, such methods directly treat relation labels as class numbers, therefore they ignore the semantics of relation labels. Recently, prompt-based fine-tuning has been proposed and attracted much attention. This kind of methods insert templates into the input and convert the classification task to a (masked) language modeling problem. With this inspiration, we propose a novel method Fine-tuning with Prompt Curriculum (FPC) for RE, with two distinctive characteristics: the relation prompt learning, introducing an auxiliary prompt-based fine-tuning task to make the model capture the semantics of relation labels; the prompt learning curriculum, a fine-tuning procedure including an increasingly difficult task to adapt the model to the difficult multitask setting. We have conducted extensive experiments on four widely used RE benchmarks under fully supervised and low-resource settings. The experimental results show that FPC can significantly outperform the existing methods and obtain the new state-of-the-art results. | FPC: Fine-tuning with Prompt Curriculum for Relation Extraction |
d235683485 | Recent pretraining models in Chinese neglect two important aspects specific to the Chinese language: glyph and pinyin, which carry significant syntax and semantic information for language understanding. In this work, we propose ChineseBERT, which incorporates both the glyph and pinyin information of Chinese characters into language model pretraining. The glyph embedding is obtained based on different fonts of a Chinese character, being able to capture character semantics from the visual features, and the pinyin embedding characterizes the pronunciation of Chinese characters, which handles the highly prevalent heteronym phenomenon in Chinese (the same character has different pronunciations with different meanings). Pretrained on large-scale unlabeled Chinese corpus, the proposed Chine-seBERT model yields significant performance boost over baseline models with fewer training steps. The proposed model achieves new SOTA performances on a wide range of Chinese NLP tasks，including machine reading comprehension, natural language inference, text classification, sentence pair matching, and competitive performances in named entity recognition and word segmentation. 1 | ChineseBERT: Chinese Pretraining Enhanced by Glyph and Pinyin Information |
d247594451 | Pre-trained models have achieved excellent performance on the dialogue task. However, for the continual increase of online chit-chat scenarios, directly fine-tuning these models for each of the new tasks not only explodes the capacity of the dialogue system on the embedded devices but also causes knowledge forgetting on pre-trained models and knowledge interference among diverse dialogue tasks. In this work, we propose a hierarchical inductive transfer framework to learn and deploy the dialogue skills continually and efficiently. First, we introduce the adapter module into pre-trained models for learning new dialogue tasks. As the only trainable module, it is beneficial for the dialogue system on the embedded devices to acquire new dialogue skills with negligible additional parameters. Then, for alleviating knowledge interference between tasks yet benefiting the regularization between them, we further design hierarchical inductive transfer that enables new tasks to use general knowledge in the base adapter without being misled by diverse knowledge in task-specific adapters. Empirical evaluation and analysis indicate that our framework obtains comparable performance under deployment-friendly model capacity. | Hierarchical Inductive Transfer for Continual Dialogue Learning |
d201668424 | Formal query generation aims to generate correct executable queries for question answering over knowledge bases (KBs), given entity and relation linking results. Current approaches build universal paraphrasing or ranking models for the whole questions, which are likely to fail in generating queries for complex, long-tail questions. In this paper, we propose SubQG, a new query generation approach based on frequent query substructures, which helps rank the existing (but nonsignificant) query structures or build new query structures. Our experiments on two benchmark datasets show that our approach significantly outperforms the existing ones, especially for complex questions. Also, it achieves promising performance with limited training data and noisy entity/relation linking results. | Leveraging Frequent Query Substructures to Generate Formal Queries for Complex Question Answering |
d219301000 | ||
d29559446 | Language identification is the task of giving a language label to a text. It is an important preprocessing step in many automatic systems operating with written text. In this paper, we present the evaluation of seven language identification methods that was done in tests between 285 languages with an out-of-domain test set. The evaluated methods are, furthermore, described using unified notation. We show that a method performing well with a small number of languages does not necessarily scale to a large number of languages. The HeLI method performs best on test lengths of over 25 characters, obtaining an F 1 -score of 99.5 already at 60 characters. | Evaluation of language identification methods using 285 languages |
d231637581 | Supervised machine translation works well when the train and test data are sampled from the same distribution. When this is not the case, adaptation techniques help ensure that the knowledge learned from out-of-domain texts generalises to in-domain sentences. We study here a related setting, multidomain adaptation, where the number of domains is potentially large and adapting separately to each domain would waste training resources. Our proposal transposes to neural machine translation the feature expansion technique of (Daumé III, 2007): it isolates domain-agnostic from domainspecific lexical representations, while sharing the most of the network across domains. Our experiments use two architectures and two language pairs: they show that our approach, while simple and computationally inexpensive, outperforms several strong baselines and delivers a multi-domain system that successfully translates texts from diverse sources. | Generic and Specialized Word Embeddings for Multi-Domain Machine Translation |
d225041022 | We present the results and findings of the First Nuanced Arabic Dialect Identification Shared Task (NADI). This Shared Task includes two subtasks: country-level dialect identification (Subtask 1) and province-level sub-dialect identification (Subtask 2). The data for the shared task covers a total of 100 provinces from 21 Arab countries and are collected from the Twitter domain. As such, NADI is the first shared task to target naturally-occurring fine-grained dialectal text at the sub-country level. A total of 61 teams from 25 countries registered to participate in the tasks, thus reflecting the interest of the community in this area. We received 47 submissions for Subtask 1 from 18 teams and 9 submissions for Subtask 2 from 9 teams. | NADI 2020: The First Nuanced Arabic Dialect Identification Shared Task |
d234742022 | While the DisCoCat model(Coecke et al., 2010)has been proved a valuable tool for studying compositional aspects of language at the level of semantics, its strong dependency on pregroup grammars poses important restrictions: first, it prevents large-scale experimentation due to the absence of a pregroup parser; and second, it limits the expressibility of the model to context-free grammars. In this paper we solve these problems by reformulating DisCoCat as a passage from Combinatory Categorial Grammar (CCG) to a category of semantics. We start by showing that standard categorial grammars can be expressed as a biclosed category, where all rules emerge as currying/uncurrying the identity; we then proceed to model permutation-inducing rules by exploiting the symmetry of the compact closed category encoding the word meaning. We provide a proof of concept for our method, converting "Alice in Wonderland" into DisCoCat form, a corpus that we make available to the community. | A CCG-Based Version of the DisCoCat Framework |
d12856343 | We use an EM algorithm to learn user models in a spoken dialog system. Our method requires automatically transcribed (with ASR) dialog corpora, plus a model of transcription errors, but does not otherwise need any manual transcription effort. We tested our method on a voice-controlled telephone directory application, and show that our learned models better replicate the true distribution of user actions than those trained by simpler methods and are very similar to user models estimated from manually transcribed dialogs. | Using Automatically Transcribed Dialogs to Learn User Models in a Spoken Dialog System |
d220045422 | While automated essay scoring (AES) can reliably grade essays at scale, automated writing evaluation (AWE) additionally provides formative feedback to guide essay revision. However, a neural AES typically does not provide useful feature representations for supporting AWE. This paper presents a method for linking AWE and neural AES, by extracting Topical Components (TCs) representing evidence from a source text using the intermediate output of attention layers. We evaluate performance using a feature-based AES requiring TCs. Results show that performance is comparable whether using automatically or manually constructed TCs for 1) representing essays as rubric-based features, 2) grading essays. | Automated Topical Component Extraction Using Neural Network Attention Scores from Source-based Essay Scoring |
d170078826 | Conversational machine comprehension (CMC) requires understanding the context of multi-turn dialogue. Using BERT, a pretraining language model, has been successful for single-turn machine comprehension, while modeling multiple turns of question answering with BERT has not been established because BERT has a limit on the number and the length of input sequences. In this paper, we propose a simple but effective method with BERT for CMC. Our method uses BERT to encode a paragraph independently conditioned with each question and each answer in a multi-turn context. Then, the method predicts an answer on the basis of the paragraph representations encoded with BERT. The experiments with representative CMC datasets, QuAC and CoQA, show that our method outperformed recently published methods (+0.8 F1 on QuAC and +2.1 F1 on CoQA). In addition, we conducted a detailed analysis of the effects of the number and types of dialogue history on the accuracy of CMC, and we found that the gold answer history, which may not be given in an actual conversation, contributed to the model performance most on both datasets. | A Simple but Effective Method to Incorporate Multi-turn Context with BERT for Conversational Machine Comprehension |
d232307330 | Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of dependency modeling among decoder inputs. In this paper, we propose CNAT, which learns implicitly categorical codes as latent variables into the non-autoregressive decoding. The interaction among these categorical codes remedies the missing dependencies and improves the model capacity. Experiment results show that our model achieves comparable or better performance in machine translation tasks than several strong baselines. | Non-Autoregressive Translation by Learning Target Categorical Codes |
d226246206 | Entity recognition provides semantic access to ancient materials in the Digital Humanities: it exposes people and places of interest in texts that cannot be read exhaustively, facilitates linking resources and can provide a window into text contents, even for texts with no translations. In this paper we present entity recognition for Coptic, the language of Hellenistic era Egypt. We evaluate NLP approaches to the task and lay out difficulties in applying them to a low-resource, morphologically complex language. We present solutions for named and non-named nested entity recognition and semi-automatic entity linking to Wikipedia, relying on robust dependency parsing, feature-based CRF models, and hand-crafted knowledge base resources, enabling high accuracy NER with orders of magnitude less data than those used for high resource languages. The results suggest avenues for research on other languages in similar settings. This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http:// creativecommons.org/licenses/by/4.0/.1 This limitation has also been recognized in NER work targeting Latin(Erdmann et al., 2016), where models based on word embeddings did not yield competitive results. | Exhaustive Entity Recognition for Coptic: Challenges and Solutions |
d214802031 | We present the first resource focusing on the verbal inflectional morphology of San Juan Quiahije Chatino, a tonal mesoamerican language spoken in Mexico. We provide a collection of complete inflection tables of 198 lemmata, with morphological tags based on the UniMorph schema. We also provide baseline results on three core NLP tasks: morphological analysis, lemmatization, and morphological inflection. | A Resource for Studying Chatino Verbal Morphology |
d222290948 | In recent years, pre-trained Transformers have dominated the majority of NLP benchmark tasks. Many variants of pre-trained Transformers have kept breaking out, and most focus on designing different pre-training objectives or variants of self-attention. Embedding the position information in the self-attention mechanism is also an indispensable factor in Transformers however is often discussed at will. Therefore, this paper carries out an empirical study on position embeddings of mainstream pre-trained Transformers, which mainly focuses on two questions: 1) Do position embeddings really learn the meaning of positions? 2) How do these different learned position embeddings affect Transformers for NLP tasks? This paper focuses on providing a new insight of pre-trained position embeddings through feature-level analysis and empirical experiments on most of iconic NLP tasks. It is believed that our experimental results can guide the future work to choose the suitable positional encoding function for specific tasks given the application property. 1 | What Do Position Embeddings Learn? An Empirical Study of Pre-Trained Language Model Positional Encoding |
d235185366 | Most language understanding models in taskoriented dialog systems are trained on a small amount of annotated training data, and evaluated in a small set from the same distribution. However, these models can lead to system failure or undesirable output when being exposed to natural language perturbation or variation in practice. In this paper, we conduct comprehensive evaluation and analysis with respect to the robustness of natural language understanding models, and introduce three important aspects related to language understanding in realworld dialog systems, namely, language variety, speech characteristics, and noise perturbation. We propose a model-agnostic toolkit LAUG to approximate natural language perturbations for testing the robustness issues in taskoriented dialog. Four data augmentation approaches covering the three aspects are assembled in LAUG, which reveals critical robustness issues in state-of-the-art models. The augmented dataset through LAUG can be used to facilitate future research on the robustness testing of language understanding in task-oriented dialog. | Robustness Testing of Language Understanding in Task-Oriented Dialog |
d257985701 | SANPSY -Univ. Bordeaux -CHU Pellegrin -CNRS -USR 3413 RÉSUMÉ. Cet article traite de la détection automatique de la somnolence dans la voix en vue de l'amélioration du suivi des patients souffrant de maladies neuropsychiatriques chroniques. Notre première approche s'inspire des systèmes état de l'art mais en les appliquant au cas particulier des patients atteints de somnolence diurne excessive (SDE). Nous basons notre étude sur le corpus TILE, qui diffère des autres corpus existants par le fait que les sujets enregistrés souffrent de SDE et que leur niveau de somnolence est mesuré de manière subjective mais aussi objective. Le système proposé permet de détecter la somnolence objective grâce à des paramètres vocaux simples et explicables à des non-spécialistes. Par la suite, nous avons développé une approche originale basée sur les erreurs de lecture que nous avons confrontées aux différentes mesures de somnolence du corpus. Nous montrons ici que relever ces erreurs peut être utile pour élaborer des marqueurs robustes de la somnolence objective.ABSTRACT. This paper deals with automatic sleepiness state estimation using speech applied to the follow-up of patients suffering from chronical neuropsychiatric diseases. Our first approach draws from state-of-the-art systems to estimate sleepiness level from voice, for the specific case of patients suffering from Excessive Daytime Sleepiness (EDS). We base our study on the MSLT corpus, that differs from other existing corpus by the fact that recorded subjects suffer from EDS and that their sleepiness level is measured by both subjective and objective means. The proposed system allows to detect objective sleepiness with simple vocal markers that are explainable to non-specialists. Furthermore, we devised a new method based on the reading errors and investigate their links with sleepiness measurements. We show that evaluating these reading errors can be useful to elaborate robust markers of objective sleepiness. MOTS-CLÉS : détection de la somnolence, marqueurs vocaux, erreurs de lecture. | Détection de la somnolence dans la voix : nouveaux marqueurs et nouvelles stratégies |
d8817648 | This paper argues that algorithms for the generation of referring expressions should aim to make it easy for hearers and readers to find the referent of the expressions that are generated. To illustrate this claim, an algorithm is described for the generation of expressions that refer across a hierarchically ordered domain, and which takes search effort into account by adding logically redundant information. To support the ideas underlying the algorithm, a psycholinguistic experiment is presented that confirms readers' preference for the generated, logically redundant expressions over non-redundant alternatives. | Generating Easy References: the Case of Document Deixis |
d6836806 | Lexico-semantic networks such as WordNet have been criticized about the nature of the senses they distinguish as well as on the way they define these senses. In this article, we present a possible solution to overcome these limits by defining the sense of words from the way they are used. More precisely, we propose to differentiate the senses of a word from a network of lexical cooccurrences built from a large corpus. This method was tested both for French and English and was evaluated for English by comparing its results with WordNet. | Discovering word senses from a network of lexical cooccurrences |
d32821791 | Temporal relation classification is becoming an active research field. Lots of methods have been proposed, while most of them focus on extracting features from external resources. Less attention has been paid to a significant advance in a closely related task: relation extraction. In this work, we borrow a state-of-the-art method in relation extraction by adopting bidirectional long short-term memory (Bi-LSTM) along dependency paths (DP). We make a "common root" assumption to extend DP representations of cross-sentence links. In the final comparison to two stateof-the-art systems on TimeBank-Dense, our model achieves comparable performance, without using external knowledge and manually annotated attributes of entities (class, tense, polarity, etc.). | Classifying Temporal Relations by Bidirectional LSTM over Dependency Paths |
d17139843 | In this paper we present our experiments in parsing Hindi. We first explored Malt and MST parsers. Considering pros of both these parsers, we developed a hybrid approach combining the output of these two parsers in an intuitive manner. We report our results on both development and test data provided in the Hindi Shared Task on Parsing at workshop on MT and parsing in Indian Languages, Coling 2012. Our system secured labeled attachment score of 90.66% and 80.77% for gold standard and automatic tracks respectively. These accuracies are 3 rd best and 5 th best for gold standard and automatic tracks respectively. | Hindi Dependency Parsing using a combined model of Malt and MST |
d234487247 | Diacritic restoration plays an important role in Natural Language Processing (NLP) for many diacritical languages such as Vietnamese, Czech, Hungarian, etc. With the development of deep neural network, this task could reach a good accuracy, i.e. F1-score is up to 97.7% for the state-of-the-art models. However, the output of these models can include meaningless syllables, the processing time is rather long especially with the sequence-tosequence method and beam search. This task is a very first step in any text (pre-)processing, which can be a part of another application. Therefore, the processing time is extremely important. To balance both accuracy and time consuming, this paper proposes a novel hybrid method which includes a transformer decoder and a diacritic penalty layer. The transformer decoder is good enough for this problem since an input character only corresponds to exact one output character. The purpose of the penalty layer is to guide the model to produce only possible diacritic letters of the language. The experimental results on Vietnamese corpus show that the proposed model helps the predicting time reduce from about eight to ten times compared to the previous methods. Whereas, the accuracy of the proposed method is better than (i.e. 1%) or equal to the state-of-the-art sequence-to-sequence without or with beam search. * | TDP -A Hybrid Diacritic Restoration with Transformer Decoder |
d218870140 | Following the major success of neural language models (LMs) such as BERT or GPT-2 on a variety of language understanding tasks, recent work focused on injecting (structured) knowledge from external resources into these models. While on the one hand, joint pretraining (i.e., training from scratch, adding objectives based on external knowledge to the primary LM objective) may be prohibitively computationally expensive, post-hoc fine-tuning on external knowledge, on the other hand, may lead to the catastrophic forgetting of distributional knowledge. In this work, we investigate models for complementing the distributional knowledge of BERT with conceptual knowledge from ConceptNet and its corresponding Open Mind Common Sense (OMCS) corpus, respectively, using adapter training. While overall results on the GLUE benchmark paint an inconclusive picture, a deeper analysis reveals that our adapter-based models substantially outperform BERT (up to 15-20 performance points) on inference tasks that require the type of conceptual knowledge explicitly present in ConceptNet and OMCS. We also open source all our experiments and relevant code under: https://github.com/ wluper/retrograph. | Common Sense or World Knowledge? Investigating Adapter-Based Knowledge Injection into Pretrained Transformers |
d12271006 | Sentiment classification aims to automatically predict sentiment polarity (e.g., positive or negative) of user generated sentiment data (e.g., reviews, blogs). Due to the mismatch among different domains, a sentiment classifier trained in one domain may not work well when directly applied to other domains. Thus, domain adaptation for sentiment classification algorithms are highly desirable to reduce the domain discrepancy and manual labeling costs. To address the above challenge, we propose a novel domain adaptation method, called Bi-Transferring Deep Neural Networks (BTDNNs). The proposed BTDNNs attempts to transfer the source domain examples to the target domain, and also transfer the target domain examples to the source domain. The linear transformation of BTDNNs ensures the feasibility of transferring between domains, and the distribution consistency between the transferred domain and the desirable domain is constrained with a linear data reconstruction manner. As a result, the transferred source domain is supervised and follows similar distribution as the target domain. Therefore, any supervised method can be used on the transferred source domain to train a classifier for sentiment classification in a target domain. We conduct experiments on a benchmark composed of reviews of 4 types of Amazon products. Experimental results show that our proposed approach significantly outperforms the several baseline methods, and achieves an accuracy which is competitive with the state-of-the-art method for domain adaptation. | Bi-Transferring Deep Neural Networks for Domain Adaptation |
d216056206 | The primary obstacle to developing technologies for low-resource languages is the lack of usable data. In this paper, we report the adoption and deployment of 4 technology-driven methods of data collection for Gondi, a low-resource vulnerable language spoken by around 2.3 million tribal people in south and central India. In the process of data collection, we also help in its revival by expanding access to information in Gondi through the creation of linguistic resources that can be used by the community, such as a dictionary, children's stories, an app with Gondi content from multiple sources and an Interactive Voice Response (IVR) based mass awareness platform. At the end of these interventions, we collected a little less than 12,000 translated words and/or sentences and identified more than 650 community members whose help can be solicited for future translation efforts. The larger goal of the project is collecting enough data in Gondi to build and deploy viable language technologies like machine translation and speech to text systems that can help take the language onto the internet. | Learnings from Technological Interventions in a Low Resource Language: A Case-Study on Gondi |
d34665650 | Solving the cold-start problem in review spam detection is an urgent and significant task. It can help the on-line review websites to relieve the damage of spammers in time, but has never been investigated by previous work. This paper proposes a novel neural network model to detect review spam for the cold-start problem, by learning to represent the new reviewers' review with jointly embedded textual and behavioral information. Experimental results prove the proposed model achieves an effective performance and possesses preferable domain-adaptability. It is also applicable to a large-scale dataset in an unsupervised way. | Handling Cold-Start Problem in Review Spam Detection by Jointly Embedding Texts and Behaviors |
d34856875 | We present neural network-based joint models for Chinese word segmentation, POS tagging and dependency parsing. Our models are the first neural approaches for fully joint Chinese analysis that is known to prevent the error propagation problem of pipeline models. Although word embeddings play a key role in dependency parsing, they cannot be applied directly to the joint task in the previous work. To address this problem, we propose embeddings of character strings, in addition to words. Experiments show that our models outperform existing systems in Chinese word segmentation and POS tagging, and perform preferable accuracies in dependency parsing. We also explore bi-LSTM models with fewer features. | Neural Joint Model for Transition-based Chinese Syntactic Analysis |
d260165143 | VerbNet est une ressource lexicale pour les verbes anglais qui est largement utilisée en TAL du fait de sa bonne couverture lexicale et syntaxique et de son encodage systématique des rôles thématiques. Aucune ressource équivalente n'existe pour le français, ce qui nous a motivés pour développer un Verb@net du français. Nous présentons comment nous avons développé Verb@net à partir de VerbNet tout en utilisant au maximum les ressources lexicales existantes du français, et comment sont encodées les différentes alternances du français en mettant l'accent sur les différences avec l'anglais (l'existence de formes pronominales, par exemple). Cet article devrait permettre à un chercheur en TAL une utilisation simple et efficace de Verb@net pour une tâche comme l'annotation en rôles sémantiques.ABSTRACT. VerbNet is a lexical resource for English verbs that has proven useful for NLP thanks to its high lexical and syntactic coverage and its systematic coding of thematic roles. Such a resource doesn't exist for French. This has motivated us to develop a Verb@net for French. We present how we have developed Verb@net from VerbNet while using as far as possible the available lexical resources for French, and how the various French alternations are coded, focusing on differences with English (existence of pronominal forms, for example). This paper should allow an NLP researcher to use Verb@net in a simple and efficient way for a task such as semantic role labeling. MOTS-CLÉS : VerbNet, rôle thématique, cadre de sous-catégorisation, alternance syntaxique. | Un Verb@net du français |
d247451124 | We present SCIENCEWORLD, a benchmark to test agents' scientific reasoning abilities in a new interactive text environment at the level of a standard elementary school science curriculum. Despite the transformer-based progress seen in question-answering and scientific text processing, we find that current models cannot reason about or explain learned science concepts in novel contexts. For instance, models can easily answer what the conductivity of a known material is but struggle when asked how they would conduct an experiment in a grounded environment to find the conductivity of an unknown material. This begs the question of whether current models are simply retrieving answers by way of seeing a large number of similar examples or if they have learned to reason about concepts in a reusable manner. We hypothesize that agents need to be grounded in interactive environments to achieve such reasoning capabilities. Our experiments provide empirical evidence supporting this hypothesisshowing that a 1.5 million parameter agent trained interactively for 100k steps outperforms a 11 billion parameter model statically trained for scientific question-answering and reasoning from millions of expert demonstrations. 12 | SCIENCEWORLD: Is your Agent Smarter than a 5 th Grader? |
d127950074 | The goal of homomorphic encryption is to encrypt data such that another party can operate on it without being explicitly exposed to the content of the original data. We introduce an idea for a privacy-preserving transformation on natural language data, inspired by homomorphic encryption. Our primary tool is obfuscation, relying on the properties of natural language. Specifically, a given English text is obfuscated using a neural model that aims to preserve the syntactic relationships of the original sentence so that the obfuscated sentence can be parsed instead of the original one. The model works at the word level, and learns to obfuscate each word separately by changing it into a new word that has a similar syntactic role. The text obfuscated by our model leads to better performance on three syntactic parsers (two dependency and one constituency parsers) in comparison to an upper-bound random substitution baseline. More specifically, the results demonstrate that as more terms are obfuscated (by their part of speech), the substitution upper bound significantly degrades, while the neural model maintains a relatively high performing parser. All of this is done without much sacrifice of privacy compared to the random substitution upper bound. We also further analyze the results, and discover that the substituted words have similar syntactic properties, but different semantic content, compared to the original words. | Obfuscation for Privacy-preserving Syntactic Parsing |
d235377459 | Large volumes of interaction logs can be collected from NLP systems that are deployed in the real world. How can this wealth of information be leveraged? Using such interaction logs in an offline reinforcement learning (RL) setting is a promising approach. However, due to the nature of NLP tasks and the constraints of production systems, a series of challenges arise. We present a concise overview of these challenges and discuss possible solutions. * All authors contributed equally, order has been randomized (see https://bit.ly/38PgRjm). | Offline Reinforcement Learning from Human Feedback in Real-World Sequence-to-Sequence Tasks |
d224817876 | We reduce the task of (span-based) PropBankstyle semantic role labeling (SRL) to syntactic dependency parsing. Our approach is motivated by our empirical analysis that shows three common syntactic patterns account for over 98% of the SRL annotations for both English and Chinese data. Based on this observation, we present a conversion scheme that packs SRL annotations into dependency tree representations through joint labels that permit highly accurate recovery back to the original format. This representation allows us to train statistical dependency parsers to tackle SRL and achieve competitive performance with the current state of the art. Our findings show the promise of syntactic dependency trees in encoding semantic role relations within their syntactic domain of locality, and point to potential further integration of syntactic methods into semantic role labeling in the future. | Semantic Role Labeling as Syntactic Dependency Parsing |
d52099556 | We propose a novel model for multi-label text classification, which is based on sequenceto-sequence learning.The model generates higher-level semantic unit representations with multi-level dilated convolution as well as a corresponding hybrid attention mechanism that extracts both the information at the word-level and the level of the semantic unit. Our designed dilated convolution effectively reduces dimension and supports an exponential expansion of receptive fields without loss of local information, and the attention-overattention mechanism is able to capture more summary relevant information from the source context. Results of our experiments show that the proposed model has significant advantages over the baseline models on the dataset RCV1-V2 and Ren-CECps, and our analysis demonstrates that our model is competitive to the deterministic hierarchical models and it is more robust to classifying low-frequency labels 1 . | Semantic-Unit-Based Dilated Convolution for Multi-Label Text Classification |
d210722425 | In a multilingual or sociolingual configuration Intra-sentential Code Switching (ICS) or Code Mixing (CM) is frequently observed nowadays. In the world most of the people know more than one language. The CM usage is especially apparent in social media platforms. Moreover, ICS is particularly significant in the context of technology, health and law where conveying the upcoming developments are difficult in one's native language. In applications like dialog systems, machine translation, semantic parsing, shallow parsing, etc. CM and Code Switching pose serious challenges. To do any further advancement in code-mixed data, the necessary step is Language Identification. So, in this paper we present a study of various models -Nave Bayes Classifier, Random Forest Classifier, Conditional Random Field (CRF) and Hidden Markov Model (HMM) for Language Identification in English -Telugu Code Mixed Data. Considering the paucity of resources in code mixed languages, we proposed CRF model and HMM model for word level language identification. Our best performing system is CRF-based with an f1-score of 0.91. 1 www.chaibisket.com | Word Level Language Identification in English Telugu Code Mixed Data |
d6162105 | OBJECTIVE:The prior knowledge about the rhetorical structure of scientific abstracts is useful for various text-mining tasks such as information extraction, information retrieval, and automatic summarization. This paper presents a novel approach to categorize sentences in scientific abstracts into four sections, objective, methods, results, and conclusions. METHOD: Formalizing the categorization task as a sequential labeling problem, we employ Conditional Random Fields (CRFs) to annotate section labels into abstract sentences. The training corpus is acquired automatically from Medline abstracts. RESULTS: The proposed method outperformed the previous approaches, achieving 95.5% per-sentence accuracy and 68.8% per-abstract accuracy. CONCLUSION: The experimental results showed that CRFs could model the rhetorical structure of abstracts more suitably. | Identifying Sections in Scientific Abstracts using Conditional Random Fields |
d202750013 | Finding the right reviewers to assess the quality of conference submissions is a time consuming process for conference organizers. Given the importance of this step, various automated reviewer-paper matching solutions have been proposed to alleviate the burden. Prior approaches, including bag-ofwords models and probabilistic topic models have been inadequate to deal with the vocabulary mismatch and partial topic overlap between a paper submission and the reviewer's expertise. Our approach, the common topic model, jointly models the topics common to the submission and the reviewer's profile while relying on abstract topic vectors. Experiments and insightful evaluations on two datasets demonstrate that the proposed method achieves consistent improvements compared to available state-of-the-art implementations of paper-reviewer matching. | PaRe: A Paper-Reviewer Matching Approach Using a Common Topic Space |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.