_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d772413 | We introduce the problem of generation in distributional semantics: Given a distributional vector representing some meaning, how can we generate the phrase that best expresses that meaning? We motivate this novel challenge on theoretical and practical grounds and propose a simple data-driven approach to the estimation of generation functions. We test this in a monolingual scenario (paraphrase generation) as well as in a cross-lingual setting (translation by synthesizing adjectivenoun phrase vectors in English and generating the equivalent expressions in Italian). | How to make words with vectors: Phrase generation in distributional semantics |
d6715557 | Information extraction techniques automatically create structured databases from unstructured data sources, such as the Web or newswire documents. Despite the successes of these systems, accuracy will always be imperfect. For many reasons, it is highly desirable to accurately estimate the confidence the system has in the correctness of each extracted field. The information extraction system we evaluate is based on a linear-chain conditional random field (CRF), a probabilistic model which has performed well on information extraction tasks because of its ability to capture arbitrary, overlapping features of the input in a Markov model. We implement several techniques to estimate the confidence of both extracted fields and entire multi-field records, obtaining an average precision of 98% for retrieving correct fields and 87% for multi-field records. | Confidence Estimation for Information Extraction |
d241583434 | Argumentation mining aims at extracting, analysing and modelling people's arguments, but large, high-quality annotated datasets are limited, and no multimodal datasets exist for this task. In this paper, we present M-Arg, a multimodal argument mining dataset with a corpus of US 2020 presidential debates, annotated through crowd-sourced annotations. This dataset allows models to be trained to extract arguments from natural dialogue such as debates using information like the intonation and rhythm of the speaker. Our dataset contains 7 hours of annotated US presidential debates, 6527 utterances and 4104 relation labels, and we report results from different baseline models, namely a text-only model, an audio-only model and multimodal models that extract features from both text and audio. With accuracy reaching 0.86 in multimodal models, we find that audio features provide added value with respect to text-only models. | M-Arg: Multimodal Argument Mining Dataset for Political Debates with Audio and Transcripts |
d33414645 | Well-nested word languages have been advertised as adequate formalizations of the notion of mild context-sensitivity. The main result of this paper is a characterization of well-nested tree languages in terms of simple attributed tree transducers. | Well-Nested Tree Languages and Attributed Tree Transducers |
d5161367 | This paper proposes a novel semisupervised word alignment technique called EMDC that integrates discriminative and generative methods. A discriminative aligner is used to find high precision partial alignments that serve as constraints for a generative aligner which implements a constrained version of the EM algorithm. Experiments on small-size Chinese and Arabic tasks show consistent improvements on AER. We also experimented with moderate-size Chinese machine translation tasks and got an average of 0.5 point improvement on BLEU scores across five standard NIST test sets and four other test sets. | EMDC: A Semi-supervised Approach for Word Alignment |
d16569528 | Statistical Machine Translation (SMT) systems are heavily dependent on the quality of parallel corpora used to train translation models. Translation quality between certain Indian languages is often poor due to the lack of training data of good quality. We used triangulation as a technique to improve the quality of translations in cases where the direct translation model did not perform satisfactorily. Triangulation uses a third language as a pivot between the source and target languages to achieve an improved and more efficient translation model in most cases. We also combined multi-pivot models using linear mixture and obtained significant improvement in BLEU scores compared to the direct source-target models. | Exploring System Combination approaches for Indo-Aryan MT Systems |
d222290667 | Language models (LMs) trained on large quantities of text have been claimed to acquire abstract linguistic representations. Our work tests the robustness of these abstractions by focusing on the ability of LMs to learn interactions between different linguistic representations. In particular, we utilized stimuli from psycholinguistic studies showing that humans can condition reference (i.e. coreference resolution) and syntactic processing on the same discourse structure (implicit causality). We compared both transformer and long short-term memory LMs to find that, contrary to humans, implicit causality only influences LM behavior for reference, not syntax, despite model representations that encode the necessary discourse information. Our results further suggest that LM behavior can contradict not only learned representations of discourse but also syntactic agreement, pointing to shortcomings of standard language modeling. | Discourse structure interacts with reference but not syntax in neural language models |
d195798924 | We address the task of assessing discourse coherence, an aspect of text quality that is essential for many NLP tasks, such as summarization and language assessment. We propose a hierarchical neural network trained in a multitask fashion that learns to predict a documentlevel coherence score (at the network's top layers) along with word-level grammatical roles (at the bottom layers), taking advantage of inductive transfer between the two tasks. We assess the extent to which our framework generalizes to different domains and prediction tasks, and demonstrate its effectiveness not only on standard binary evaluation coherence tasks, but also on real-world tasks involving the prediction of varying degrees of coherence, achieving a new state of the art. | Multi-Task Learning for Coherence Modeling |
d250164464 | Text simplification is a method for improving the accessibility of text by converting complex sentences into simple sentences. Multiple studies have been done to create datasets for text simplification. However, most of these datasets focus on high-resource languages only. In this work, we proposed a complex word dataset for Hindi, a language largely ignored in text simplification literature. We used various Hindi knowledge annotators for annotation to capture the annotator's language knowledge. Our analysis shows a significant difference between native and non-native annotators' perception of word complexity. We also built an automatic complex word classifier using a soft voting approach based on the predictions from tree-based ensemble classifiers. These models behave differently for annotations made by different categories of users, such as native and non-native speakers. Our dataset and analysis will help simplify Hindi text depending on the user's language understanding. The dataset is available at https://zenodo.org/record/5229160. | CWID-hi: A Dataset for Complex Word Identification in Hindi Text |
d233365156 | We present three methods developed for the Shared Task on Sarcasm and Sentiment Detection in Arabic. We present a baseline that uses character n-gram features. We also propose two more sophisticated methods: a recurrent neural network with a word level representation and an ensemble classifier relying on word and character-level features. We chose to present results from an ensemble classifier but it was not very successful as compared to the best systems : 22th/37 on sarcasm detection and 15th/22 on sentiment detection. It finally appeared that our baseline could have easily been tuned and achieve much better results. | Sarcasm and Sentiment Detection in Arabic: Investigating the Interest of Character-level Features |
d220046855 | Definition generation, which aims to automatically generate dictionary definitions for words, has recently been proposed to assist the construction of dictionaries and help people understand unfamiliar texts. However, previous works hardly consider explicitly modeling the "components" of definitions, leading to under-specific generation results. In this paper, we propose ESD, namely Explicit Semantic Decomposition for definition generation, which explicitly decomposes meaning of words into semantic components, and models them with discrete latent variables for definition generation. Experimental results show that ESD achieves substantial improvements on WordNet and Oxford benchmarks over strong previous baselines. | Explicit Semantic Decomposition for Definition Generation |
d226283865 | Automated agents ("bots") have emerged as an ubiquitous and influential presence on social media. Bots engage on social media platforms by posting content and replying to other users on the platform. In this work we conduct an empirical analysis of the activity of a single bot on Reddit. Our goal is to determine whether bot activity (in the form of posted comments on the website) has an effect on how humans engage on Reddit. We find that (1) the sentiment of a bot comment has a significant, positive effect on the subsequent human reply, and (2) human Reddit users modify their comment behaviors to overlap with the text of the bot, similar to how humans modify their text to mimic other humans in conversation. Understanding human-bot interactions on social media with relatively simple bots is important for preparing for more advanced bots in the future. | An Empirical Analysis of Human-Bot Interaction on Reddit |
d59524257 | 摘要 可讀性模型可以自動估測文本的可讀性，幫助讀者去挑選符合自己閱讀能力的文件，以 達到更好的學習效果。長久以來，研究人員致力於可讀性模型或特徵的研發，尤其近年 來隨著表示學習法的蓬勃發展，使得訓練可讀性模型所需要的特徵可以不再需要仰賴專 家，這也使得可讀性模型的發展有了一個嶄新的研究方向。然而，不同表示學習法所抽 取出來的特徵各有所長，但過去的可讀性研究大多只用單一方法來訓練可讀性模型。因 此，本論文嘗試利用類神經網路來融合卷積神經網路及快速文本兩種表示學習法，以訓 練出一個能夠分析跨領域文件的可讀性模型，並可以因應文件內容多元主題的特性。從 實驗結果可以發現本論文所提出的可讀性模型，其效能可略微勝出單一表示學習法所訓 練的可讀性模型。 關鍵詞：可讀性，詞向量，卷積神經網路，表示學習法，快速文本 | Exploring Combination of FastText and Convolutional Neural Networks for Building Readability Models |
d10338752 | This paper presents the syntactic and semantic tags used to annotate predicate-argument structure in the Berkeley FrameNet Project. It briefly explains the theory of frame semantics on which semantic annotation is based, discusses possible applications of FrameNet annotation, and compares FrameNet to other prominent iexical resources. | The FrameNet tagset for frame-semantic and syntactic coding of predicate-argument structure |
d21703891 | In this paper, we introduce a two-step corpora-based methodology, starting from a corpus of human-human interactions to construct a semi-autonomous system in order to collect a new corpus of human-machine interaction, a step before the development of a fully autonomous system constructed based on the analysis of the collected corpora. The presented methodology is illustrated in the context of a virtual reality training platform for doctors breaking bad news. | A Semi-autonomous System for Creating a Human-Machine Interaction Corpus in Virtual Reality: Application to the ACORFORMed System for Training Doctors to Break Bad News |
d227230689 | We reduce the model size of pre-trained word embeddings by a factor of 200 while preserving its quality. Previous studies in this direction created a smaller word embedding model by reconstructing pre-trained word representations from those of subwords, which allows to store only a smaller number of subword embeddings in the memory. However, previous studies that train the reconstruction models using only target words cannot reduce the model size extremely while preserving its quality. Inspired by the observation of words with similar meanings having similar embeddings, our reconstruction training learns the global relationships among words, which can be employed in various models for word embedding reconstruction. Experimental results on word similarity benchmarks show that the proposed method improves the performance of the all subword-based reconstruction models. | Tiny Word Embeddings Using Globally Informed Reconstruction |
d53473481 | We introduce CALIMA Star , a very rich Arabic morphological analyzer and generator that provides functional and form-based morphological features as well as built-in tokenization, phonological representation, lexical rationality and much more. This tool includes a fast engine that can be easily integrated into other systems, as well as an easy-to-use API and a web interface. CALIMA Star also supports morphological reinflection. We evaluate CALIMA Star against four commonly used analyzers for Arabic in terms of speed and morphological content. | An Arabic Morphological Analyzer and Generator with Copious Features |
d252624600 | Online news consumption plays an important role in shaping the political opinions of citizens. The news is often served by recommendation algorithms, which adapt content to users' preferences. Such algorithms can lead to political polarization as the societal effects of the recommended content and recommendation design are disregarded. We posit that biases appear, at least in part, due to a weak entanglement between natural language processing and recommender systems, both processes yet at work in the diffusion and personalization of online information. We assume that both diversity and acceptability of recommended content would benefit from such a synergy. We discuss the limitations of current approaches as well as promising leads of opinion-mining integration for the political news recommendation process. | Don't Burst Blindly: For a Better Use of Natural Language Processing to Fight Opinion Bubbles in News Recommendations |
d9691480 | We introduce a notion of training methodology space (TM space) for specifying training methodologies in tile different disciplines and teaching traditions associated with computational linguistics and the human language technologies, and pin our approach to the concept of operational model; we also discuss different general levels of interactivity. A number of operational models are introduced, with web interfaces for lexical databases, DFSA matrices, finitestate phonotactics development, and DATR lexica. | Web tools for introductory computational linguistics |
d219307723 | In this paper we describe the SemEval-2010 Cross-Lingual Lexical Substitution task, where given an English target word in context, participating systems had to find an alternative substitute word or phrase in Spanish. The task is based on the English Lexical Substitution task run at SemEval-2007. In this paper we provide background and motivation for the task, we describe the data annotation process and the scoring system, and present the results of the participating systems. | SemEval-2010 Task 2: Cross-Lingual Lexical Substitution |
d12781220 | Organizing data into category hierarchies (taxonomies) is useful for content discovery, search, exploration and analysis. In industrial settings targeted taxonomies for specific domains are mostly created manually, typically by domain experts, which is time consuming and requires a high level of expertise. This paper presents an algorithm and an implemented interactive system for automatically generating target-domain taxonomies based on the Wikipedia Category Hierarchy. The system also enables human post-editing, facilitated by intelligent assistance. | A Support Tool for Deriving Domain Taxonomies from Wikipedia |
d17089657 | It is widely accepted that translating usergenerated (UG) text is a difficult task for modern statistical machine translation (SMT) systems. The translation quality metrics typically used in the SMT literature reflect the overall quality of the system output but provide little insight into what exactly makes UG text translation difficult. This paper analyzes in detail the behavior of a state-of-the-art SMT system on five different types of informal text. The results help to demystify the poor SMT performance experienced by researchers who use SMT as an intermediate step of their UG-NLP pipeline, and to identify translation modeling aspects that the SMT community should more urgently address to improve translation of UG data. | Five Shades of Noise: Analyzing Machine Translation Errors in User-Generated Text |
d227231247 | Models with a large number of parameters are prone to over-fitting and often fail to capture the underlying input distribution. We introduce Emix, a data augmentation method that uses interpolations of word embeddings and hidden layer representations to construct virtual examples. We show that Emix shows significant improvements over previously used interpolation based regularizers and data augmentation techniques. We also demonstrate how our proposed method is more robust to sparsification. We highlight the merits of our proposed methodology by performing thorough quantitative and qualitative assessments.This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http:// creativecommons.org/licenses/by/4.0/. | Augmenting NLP models using Latent Feature Interpolations |
d202763066 | Discourse parsing could not yet take full advantage of the neural NLP revolution, mostly due to the lack of annotated datasets. We propose a novel approach that uses distant supervision on an auxiliary task (sentiment classification), to generate abundant data for RSTstyle discourse structure prediction. Our approach combines a neural variant of multipleinstance learning, using document-level supervision, with an optimal CKY-style tree generation algorithm. In a series of experiments, we train a discourse parser (for only structure prediction) on our automatically generated dataset and compare it with parsers trained on human-annotated corpora (news domain RST-DT and Instructional domain). Results indicate that while our parser does not yet match the performance of a parser trained and tested on the same dataset (intra-domain), it does perform remarkably well on the much more difficult and arguably more useful task of inter-domain discourse structure prediction, where the parser is trained on one domain and tested/applied on another one. | Predicting Discourse Structure using Distant Supervision from Sentiment |
d2703746 | In the development of a machine translation system, one important issue is being able to adapt to a specific domain without requiring timeconsuming lexical work.We have experimented with using a statistical word-alignment algorithm to derive word association pairs (French-English) that complement an existing multipurpose bilingual dictionary. This word association information is added to the system at the time of the automatic creation of our translation pattern database, thereby making this database more domain specific. This technique significantly improves the overall quality of translation, as measured in an independent blind evaluation. French training sentence: Dans le menu Démarrer, pointez sur Programmes, sur Outils d'administration (commun), puis cliquez sur Gestionnaire des utilisateurs pour les domaines. English training sentence: On the Start menu, point to Programs, point to Administrative Tools (Common), and then click User Manager for Domains.The French-English lexicon is used during the training period of the transfer component to establish initial, tentative, word correspondences during the alignment process. The sources for the bilingual dictionary were: Cambridge University Press English-French, Soft-Art | Adding Domain Specificity to an MT system |
d249538120 | We present a Multilingual Open Text (MOT), a new multilingual corpus containing text in 44 languages, many of which have limited existing text resources for natural language processing. The first release of the corpus contains over 2.8 million news articles and an additional 1 million short snippets (photo captions, video descriptions, etc.) published between 2001-2022 and collected from Voice of America's news websites. We describe our process for collecting, filtering, and processing the data. The source material is in the public domain, our collection is licensed using a creative commons license (CC BY 4.0), and all software used to create the corpus is released under the MIT License. The corpus will be regularly updated as additional documents are published. | Multilingual Open Text Release 1: Public Domain News in 44 Languages |
d30648313 | This paper presents doctoral thesis in progress of formation of corpus-sample Coelho Netto, a corpus compiled, annotated and analyzed morphosyntactically occurrences and opulence of verbs, adjectives and adverbs in-ly compared the texts of the same period, as well as observation the accuracy of automatic labeler Aelius (free software in Python), model Hunpos, trained at Corpus of Historical Portuguese Tycho Brahe (CHPTB) in literary texts by this author: two novels and three short stories of the period between the end of the 19th century and early 20th century. The Corpus is approximately forty-seven thousand tokens (words and punctuations). Resumo. Este trabalho apresenta a tese de doutoramento em andamento da constituição do corpus-amostra Coelho Netto, um corpus compilado, anotado morfossintaticamente e analisadas as ocorrências e opulência de verbos, adjetivos e advérbios em -mente em comparação a textos do mesmo período, além de observação da acurácia do etiquetador automático Aelius (software livre em Python), modelo Hunpos, treinado no Corpus Histórico do Português Tycho Brahe (CHPTB) em textos literários desse autor: dois romances e três contos do período entre final do século XIX e início do século XX. O Corpus tem aproximadamente quarenta e sete mil tokens (palavras e pontuações). | Corpus-amostra Coelho Netto: compilação, anotação e ocorrências em textos literários dos séc. XIX e XX |
d52009101 | Machine reading comprehension (MRC) has recently attracted attention in the fields of natural language processing and machine learning. One of the problematic presumptions with current MRC technologies is that each question is assumed to be answerable by looking at a given text passage. However, to realize human-like language comprehension ability, a machine should also be able to distinguish not-answerable questions (NAQs) from answerable questions. To develop this functionality, a dataset incorporating hard-to-detect NAQs is vital; however, its manual construction would be expensive. This paper proposes a dataset creation method that alters an existing MRC dataset, the Stanford Question Answering Dataset, and describes the resulting dataset. The value of this dataset is likely to increase if each NAQ in the dataset is properly classified with the difficulty of identifying it as an NAQ. This difficulty level would allow researchers to evaluate a machine's NAQ detection performance more precisely. Therefore, we propose a method for automatically assigning difficulty level labels, which basically measures the similarity between a question and the target text passage. Our NAQ detection experiments demonstrate that the resulting dataset, having difficulty level annotations, is valid and potentially useful in the development of advanced MRC models. This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http: //creativecommons.org/licenses/by/4.0/ | Answerable or Not: Devising a Dataset for Extending Machine Reading Comprehension |
d7996527 | In this paper, we examine the benefit of applying text segmentation methods to perform language identification in forums. The focus here is on forums containing a mixture of information written in Greek, English as well as Greeklish. Greeklish can be defined as the use of Latin alphabet for rendering Greek words with Latin characters. For the evaluation, a corpus was manually created by collecting web pages from Greek university forums and most specifically, pages containing information that combines Greek with English technical terminology and Greeklish. The evaluation using two well known text segmentation algorithms leads to the conclusion that despite the difficulty of the problem examined, text segmentation seems to be a promising solution. | Text segmentation for Language Identification in Greek Forums |
d1365692 | For speaker-independent speech recognition, speaker variation is one of the major error sources. In this paper, a speaker-independentnormalization network is constructed such that speaker variation effects can be minimized. To achieve this goal, multiple speaker clusters are constructed from the speaker-independent training database. A codeword-dependent neural network is associated with each speaker cluster. The cluster that contains the largest number of speakers is designated as the golden cluster. The objective function is to minimize distortions between acoustic data in each cluster and the golden speakercluster. Performanceevaluation showedthat speakernormalized front-end reduced the error rate by 15% for the DARPA resource management speaker-independent speech recognition task. | Minimizing Speaker Variation Effects for Speaker-Independent Speech Recognition |
d2946085 | A strategic challenge for Europe in today's globalised economy is to overcome language barriers through technological means. In particular, machine translation (MT) systems are expected to have a significant impact on the management of multilingualism in Europe.PANACEA addresses the most critical aspect for MT: the so-called language resource (LR) bottleneck. Although MT technologies may consist of language-independent engines, they depend on the availability of language-dependent knowledge for their real-life implementation, i.e., they require LRs.The objective of PANACEA is to build a factory of LRs that automates the stages involved in the acquisition, production, updating and maintenance of LRs required by MT systems and by other applications based on language technologies, and simplifies eventual issues regarding intellectual property rights. This automation will cut down the cost, time and human effort significantly. These reductions of costs and time are the only way to guarantee the continuous supply of LRs that MT and other language technologies will be demanding in the multilingual Europe.In its first year, PANACEA has developed the first version of the LR factory, which allows the final user to acquire LRs by designing workflows (acquisition pipelines) that connect webservices provided by the different partners. For example, to create a parallel corpus for a given domain, the user can call the multilingual crawler (provided by ILSP) with a set of keywords of the domain, and connect its output to any of the sentential aligners (provided by DCU). | PANACEA (Platform for Automatic, Normalised Annotation and Cost-Effective Acquisition of Language Resources for Human Lan- guage Technologies) |
d10026482 | In modern machine translation practice, a statistical phrasal or hierarchical translation system usually relies on a huge set of translation rules extracted from bi-lingual training data. This approach not only results in space and efficiency issues, but also suffers from the sparse data problem. In this paper, we propose to use factorized grammars, an idea widely accepted in the field of linguistic grammar construction, to generalize translation rules, so as to solve these two problems. We designed a method to take advantage of the XTAG English Grammar to facilitate the extraction of factorized rules. We experimented on various setups of low-resource language translation, and showed consistent significant improvement in BLEU over state-ofthe-art string-to-dependency baseline systems with 200K words of bi-lingual training data. | Statistical Machine Translation with a Factorized Grammar |
d16989361 | The paper describes a supervised approach for the detection of the most frequent senses of words on the basis of RuThes thesaurus, which is a large linguistic ontology for Russian. Due to the large number of monosemous multiword expressions and the set of RuThes relations it is possible to calculate several context features for ambiguous words and to study their contribution to a supervised model for detecting frequent senses. | Determining the Most Frequent Senses Using Russian Linguistic Ontology RuThes |
d182408517 | This paper demonstrates a state-of-the-art endto-end multilingual (English, Russian, and Ukrainian) knowledge extraction system that can perform entity discovery and linking, relation extraction, event extraction, and coreference. It extracts and aggregates knowledge elements across multiple languages and documents as well as provides visualizations of the results along three dimensions: temporal (as displayed in an event timeline), spatial (as displayed in an event heatmap), and relational (as displayed in entity-relation networks). For our system to further support users' analyses of causal sequences of events in complex situations, we also integrate a wide range of human moral value measures, independently derived from region-based survey, into the event heatmap. This system is publicly available as a docker container and a live demo, 1,2 with a video demonstrating the system 3 . | Multilingual Entity, Relation, Event and Human Value Extraction |
d14920006 | We propose a statistical frame-based approach (FBA) for natural language processing, and demonstrate its advantage over traditional machine learning methods by using topic detection as a case study. FBA perceives and identifies semantic knowledge in a more general manner by collecting important linguistic patterns within documents through a unique flexible matching scheme that allows word insertion, deletion and substitution (IDS) to capture linguistic structures within the text. In addition, FBA can also overcome major issues of the rule-based approach by reducing human effort through its highly automated pattern generation and summarization. Using Yahoo! Chinese news corpus containing about 140,000 news articles, we provide a comprehensive performance evaluation that demonstrates the effectiveness of FBA in detecting the topic of a document by exploiting the semantic association and the context within the text. Moreover, it outperforms common topic models like Naïve Bayes, Vector Space Model, and LDA-SVM.On the other hand, there are several machine learning-based approaches. For instance,Nallapati et al. (2004)attempted to find characteristics of topics by clustering keywords using statistical similarity. The clusters are then connected chronologically to form a time-line of the topic. Furthermore, many previous methods treated topic detection as a supervised classification problem(Blei et al., 2003;Zhang and Wang, 2010). These approaches can achieve substantial performance without much human involvement. However, to manifest topic as- | Semantic Frame-based Statistical Approach for Topic Detection |
d15351419 | Automatic resolution of Crossword Puzzles (CPs) heavily depends on the quality of the answer candidate lists produced by a retrieval system for each clue of the puzzle grid. Previous work has shown that such lists can be generated using Information Retrieval (IR) search algorithms applied to the databases containing previously solved CPs and reranked with tree kernels (TKs) applied to a syntactic tree representation of the clues. In this paper, we create a labelled dataset of 2 million clues on which we apply an innovative Distributional Neural Network (DNN) for reranking clue pairs. Our DNN is computationally efficient and can thus take advantage of such large datasets showing a large improvement over the TK approach, when the latter uses small training data. In contrast, when data is scarce, TKs outperform DNNs. | Distributional Neural Networks for Automatic Resolution of Crossword Puzzles |
d15319313 | We describe the work carried out by the DCU-ADAPT team on the Lexical Normalisation shared task at W-NUT 2015. We train a generalised perceptron to annotate noisy text with edit operations that normalise the text when executed. Features are character n-grams, recurrent neural network language model hidden layer activations, character class and eligibility for editing according to the task rules. We combine predictions from 25 models trained on subsets of the training data by selecting the most-likely normalisation according to a character language model. We compare the use of a generalised perceptron to the use of conditional random fields restricted to smaller amounts of training data due to memory constraints. Furthermore, we make a first attempt to verify Chrupała (2014)'s hypothesis that the noisy channel model would not be useful due to the limited amount of training data for the source language model, i.e. the language model on normalised text. | DCU-ADAPT: Learning Edit Operations for Microblog Normalisation with the Generalised Perceptron |
d196019 | Data Extraction as Text Categorization : An Experiment With the MUC-3 Corpu s INTRODUCTIO N | |
d236477764 | Stereotypes are inferences drawn about people based on their demographic attributes, which may result in harms to users when a system is deployed. In generative language-inference tasks, given a premise, a model produces plausible hypotheses that follow either logically (natural language inference) or commonsensically (commonsense inference). Such tasks are therefore a fruitful setting in which to explore the degree to which NLP systems encode stereotypes. In our work, we study how stereotypes manifest when the potential targets of stereotypes are situated in real-life, neutral contexts. We collect human judgments on the presence of stereotypes in generated inferences, and compare how perceptions of stereotypes vary due to annotator positionality. | Analyzing Stereotypes in Generative Text Inference Tasks |
d18037181 | We describe our submission system to the SemEval-2016 Task 8 on Abstract Meaning Representation (AMR) Parsing. We attempt to improve AMR parsing by exploiting preposition semantic role labeling information retrieved from a multi-layer feed-forward neural network. Prepositional semantics is included as features into the transition-based AMR parsing system CAMR (Wang, Xue, and S. Pradhan 2015a). The inclusion of the features modifies the behavior of CAMR when creating meaning representations triggered by prepositional semantics. Despite the usefulness of preposition semantic role labeling information for AMR parsing, it does not have an impact to the parsing F-score of CAMR, but reduces the parsing recall by 1%. | ICL-HD at SemEval-2016 Task 8: Meaning Representation Parsing - Augmenting AMR Parsing with a Preposition Semantic Role Labeling Neural Network |
d257258164 | This article presents a comparative analysis of four different syntactic typological approaches applied to 20 different languages. We compared three specific quantitative methods, using parallel CoNLL-U corpora, to the classification obtained via syntactic features provided by a typological database (lang2vec). First, we analyzed the Marsagram linear approach which consists of extracting the frequency word-order patterns regarding the position of components inside syntactic nodes. The second approach considers the relative position of heads and dependents, and the third is based simply on the relative position of verbs and objects. From the results, it was possible to observe that each method provides different language clusters which can be compared to the classic genealogical classification (the lang2vec and the head and dependent methods being the closest). As different word-order phenomena are considered in these specific typological strategies, each one provides a different angle of analysis to be applied according to the precise needs of the researchers. | Analysis of Corpus-based Word-Order Typological Methods |
d248085462 | Pre-trained language models (PLM) are effective components of few-shot named entity recognition (NER) approaches when augmented with continued pre-training on taskspecific out-of-domain data or fine-tuning on in-domain data. However, their performance in low-resource scenarios, where such data is not available, remains an open question. We introduce an encoder evaluation framework, and use it to systematically compare the performance of state-of-the-art pre-trained representations on the task of low-resource NER. We analyze a wide range of encoders pre-trained with different strategies, model architectures, intermediate-task fine-tuning, and contrastive learning. Our experimental results across ten benchmark NER datasets in English and German show that encoder performance varies significantly, suggesting that the choice of encoder for a specific low-resource scenario needs to be carefully evaluated. | A Comparative Study of Pre-trained Encoders for Low-Resource Named Entity Recognition |
d1814363 | Traditionally, Japanese is said to have two kinds of negative word nai, meaning 'not' in English: the nai occurring after a verb or adjective and the nai with an independent status as a word. In contrast to traditional Japanese linguistics, most generative studies of Japanese have made no such distinction but rather focused their attention on such issues as the interaction between a scope of negation and a quantifier and the behavior of negative polarity items. In this paper we attempt to integrate all the proposals of each position by making use of both syntactic and semantic features. We also claim that the independent nai has a distinctly adjectival status with regard to its* semantic structure. | THE TWO KINDS OF JAPANESE NEGATIVE Nai IN TERMS OF THEIR NPI LICENSING CONDITIONS |
d5855331 | This paper describes an emerging shared repository of large-text resources for creating word vectors, including pre-processed corpora and pre-trained vectors for a range of frameworks and configurations. This will facilitate reuse, rapid experimentation, and replicability of results. | Word vectors, reuse, and replicability: Towards a community repository of large-text resources |
d203693213 | Historical cryptology is the study of historical encrypted messages aiming at their decryption by analyzing the mathematical, linguistic and other coding patterns and their historical context. In libraries and archives we can find quite a lot of ciphers, as well as keys describing the method used to transform the plaintext message into a ciphertext. In this paper, we present work on automatically mapping keys to ciphers to reconstruct the original plaintext message, and use language models generated from historical texts to guess the underlying plaintext language. | Matching Keys and Encrypted Manuscripts |
d252624703 | We introduce how the proprietary machine learning algorithms developed by Gojob, an HR Tech company, to match candidates to a job offer are as transparent and explainable as possible to users (i.e., our recruiters) and our clients (e.g. companies looking to fill jobs). We detail how our matching algorithm (which identifies the best candidates for a job offer) controls the fairness of its outcome. We have described the steps we have taken to ensure that the decisions made by our mathematical models not only inform but improve the performance of our recruiters. | Transparency and Explainability of a Machine Learning Model in the Context of Human Resource Management |
d10526118 | In this paper, a new methodology for speech corpora definition from internet documents is described, in order to record a large speech database, dedicated to the training and testing of acoustic models for speech recognition. In the first section, the Web robot which is in charge of collecting Web pages from Internet is presented, then the web text to French sentences filtering mechanism is explained. Some information about the corpus organization (90% for training and 10% for test) is given. In the third section, the phoneme distribution of the corpus is presented and comparison is made with others French language studies. Finally tools and planning for recording the speech database with more than one hundred speakers are described. | A NEW METHODOLOGY FOR SPEECH CORPORA DEFINITION FROM INTERNET DOCUMENTS |
d252819211 | Recursive processing is considered a hallmark of human linguistic abilities. A recent study evaluated recursive processing in recurrent neural language models (RNN-LMs) and showed that such models perform below chance level on embedded dependencies within nested constructions -a prototypical example of recursion in natural language. Here, we study if state-of-the-art Transformer LMs do any better. We test eight different Transformer LMs on two different types of nested constructions, which differ in whether the embedded (inner) dependency is short or long range. We find that Transformers achieve near-perfect performance on short-range embedded dependencies, significantly better than previous results reported for RNN-LMs and humans. However, on longrange embedded dependencies, Transformers' performance sharply drops below chance level.Remarkably, the addition of only three words to the embedded dependency caused Transformers to fall from near-perfect to below-chance performance. Taken together, our results reveal how brittle syntactic processing is in Transformers, compared to humans. | Can Transformers Process Recursive Nested Constructions, Like Humans? |
d237012292 | We present MT-TELESCOPE, a visualization platform designed to facilitate comparative analysis of the output quality of two Machine Translation (MT) systems. While automated MT evaluation metrics are commonly used to evaluate MT systems at a corpus-level, our platform supports fine-grained segment-level analysis and interactive visualisations that expose the fundamental differences in the performance of the compared systems. MT-TELESCOPE also supports dynamic corpus filtering to enable focused analysis on specific phenomena such as; translation of named entities, handling of terminology, and the impact of input segment length on translation quality. Furthermore, the platform provides a bootstrapped t-test for statistical significance as a means of evaluating the rigor of the resulting system ranking. MT-TELESCOPE is open source 1 , written in Python, and is built around a user friendly and dynamic web interface. Complementing other existing tools, our platform is designed to facilitate and promote the broader adoption of more rigorous analysis practices in the evaluation of MT quality. | MT-TELESCOPE An interactive platform for contrastive evaluation of MT systems |
d18324496 | The N400 is a human neuroelectric response to semantic incongruity in on-line sentence processing, and implausibility in context has been identified as one of the factors that influence the size of the N400. In this paper we investigate whether predictors derived from Latent Semantic Analysis, language models, and Roark's parser are significant in modeling of the N400m (the neuromagnetic version of the N400). We also investigate significance of a novel pairwise-priming language model based on the IBM Model 1 translation model. Our experiments show that all the predictors are significant. Moreover, we show that predictors based on the 4-gram language model and the pairwise-priming language model are highly correlated with the manual annotation of contextual plausibility, suggesting that these predictors are capable of playing the same role as the manual annotations in prediction of the N400m response. We also show that the proposed predictors can be grouped into two clusters of significant predictors, suggesting that each cluster is capturing a different characteristic of the N400m response. | Using Language Models and Latent Semantic Analysis to Characterise the N400m Neural Response |
d23661440 | Rule-based techniques and tools to extract entities and relational entities from documents allow users to specify desired entities using natural language questions, finite state automata, regular expressions, structured query language statements, or proprietary scripts. These techniques and tools require expertise in linguistics and programming and lack support of Arabic morphological analysis which is key to process Arabic text. In this work, we present MERF; a morphology-based entity and relational entity extraction framework for Arabic text. MERF provides a user-friendly interface where the user, with basic knowledge of linguistic features and regular expressions, defines tag types and interactively associates them with regular expressions defined over Boolean formulae. Boolean formulae range over matches of Arabic morphological features, and synonymity features. Users define user defined relations with tuples of subexpression matches and can associate code actions with subexpressions. MERF computes feature matches, regular expression matches, and constructs entities and relational entities from user defined relations. We evaluated our work with several case studies and compared with existing application-specific techniques. The results show that MERF requires shorter development time and effort compared to existing techniques and produces reasonably accurate results within a reasonable overhead in run time. | MERF: Morphology-based Entity and Relational Entity Extraction Framework for Arabic |
d14777142 | Word clustering which generalizes specific features cluster words in the same syntactic or semantic categories into a group. It is an effective approach to reduce feature dimensionality and feature sparseness which are clearly useful for many NLP applications. This paper proposes an unsupervised label propagation algorithm (Un-LP) for word clustering which uses multi-exemplars to represent a cluster. Experiments on a synthetic 2D dataset show the strong ability of selfcorrecting of the proposed algorithm. Besides, the experimental results on 20NG demonstrate that our algorithm outperforms the conventional cluster algorithms.Problem setupAssume that we have a corpus with N documents denoted by D = {d 1 , d 2 , · · · , d N }; each document in the corpus consists of a list of words denoted by d i = {w 1 , w 2 , · · · , w N d } where each w i is an item from a vocabulary index with V distinct terms denoted by W = {v 1 , v 2 , · · · , v V } and N d is the document This work is licensed under a Creative Commons Attribution 4.0 International License. Page numbers and proceedings footer are added by the organizers. License details: http://creativecommons.org/licenses/by/4.0/. | Word Clustering Based on Un-LP Algorithm |
d126169759 | Aspect-based sentiment analysis involves the recognition of so called opinion target expressions (OTEs). To automatically extract OTEs, supervised learning algorithms are usually employed which are trained on manually annotated corpora. The creation of these corpora is labor-intensive and sufficiently large datasets are therefore usually only available for a very narrow selection of languages and domains. In this work, we address the lack of available annotated data for specific languages by proposing a zero-shot cross-lingual approach for the extraction of opinion target expressions. We leverage multilingual word embeddings that share a common vector space across various languages and incorporate these into a convolutional neural network architecture for OTE extraction. Our experiments with 5 languages give promising results: We can successfully train a model on annotated data of a source language and perform accurate prediction on a target language without ever using any annotated samples in that target language. Depending on the source and target language pairs, we reach performances in a zero-shot regime of up to 77% of a model trained on target language data. Furthermore, we can increase this performance up to 87% of a baseline model trained on target language data by performing cross-lingual learning from multiple source languages. | Zero-Shot Cross-Lingual Opinion Target Extraction |
d150567 | We developed a method for automatically distinguishing the machine-translatable and non-machine-translatable parts of a given sentence for a particular machine translation (MT) system. They can be distinguished by calculating the similarity between a source-language sentence and its back translation for each part of the sentence. The parts with low similarities are highly likely to be non-machinetranslatable parts. We showed that the parts of a sentence that are automatically distinguished as non-machine-translatable provide useful information for paraphrasing or revising the sentence in the source language to improve the quality of the translation by the MT system. We also developed a method of providing knowledge useful to effectively paraphrasing or revising the detected non-machine-translatable parts. Two types of knowledge were extracted from the EDR dictionary: one for transforming a lexical entry into an expression used in the definition and the other for conducting the reverse paraphrasing, which transforms an expression found in a definition into the lexical entry. We found that the information provided by the methods helped improve the machine translatability of the originally input sentences. | Automatic Detection and Semi-Automatic Revision of Non-Machine-Translatable Parts of a Sentence |
d44162573 | In this paper we present our contribution to SemEval-2018, a classifier for classifying multi-label emotions of Arabic and English tweets. We attempted "Affect in Tweets", specifically Task E-c: Detecting Emotions (multi-label classification). Our method is based on preprocessing the tweets and creating word vectors combined with a self correction step to remove noise. We also make use of emotion specific thresholds. The final submission was selected upon the best performance achieved, selected when using a range of thresholds. Our system was evaluated on the Arabic and English datasets provided for the task by the competition organisers, where it ranked 2nd for the Arabic dataset (out of 14 entries) and 12th for the English dataset (out of 35 entries). | CENTEMENT at SemEval-2018 Task 1: Classification of Tweets using Multiple Thresholds with Self-correction and Weighted Conditional Probabilities |
d7124715 | Machine learning approaches to coreference resolution are typically supervised, and require expensive labeled data. Some unsupervised approaches have been proposed (e.g., Haghighi and Klein(2007)), but they are less accurate. In this paper, we present the first unsupervised approach that is competitive with supervised ones. This is made possible by performing joint inference across mentions, in contrast to the pairwise classification typically used in supervised methods, and by using Markov logic as a representation language, which enables us to easily express relations like apposition and predicate nominals. On MUC and ACE datasets, our model outperforms Haghigi and Klein's one using only a fraction of the training data, and often matches or exceeds the accuracy of state-of-the-art supervised models. | Joint Unsupervised Coreference Resolution with Markov Logic |
d1043632 | Certain spans of utterances in a discourse, referred to here as segments, are widely assumedto form coherent units. Further, the segmental structure of discourse has been claimed to constrain and be constrained by many phenomena. However, there is weak consensus on the nature of segments and the criteria for recognizing or generating them. We present quantitative results of a two part study using a corpus of spontaneous, narrative monologues. The first part evaluates the statistical reliability of human segmentation of our corpus, where speaker intention is the segmentation criterion. We then use the subjects' segmentations to evaluate the correlation of discourse segmentation with three linguistic cues (referential noun phrases, cue words, and pauses), using information retrieval metrics. | INTENTION-BASED SEGMENTATION: HUMAN RELIABILITY AND CORRELATION WITH LINGUISTIC CUES |
d17388420 | ON THE ACQUISITION OF CONCEPTUAL DEFINITIONS VIA TEXTUAL MODELLING OF MEANING PARAPHRASES | |
d256460961 | This paper describes insights into how different inference algorithms structure discourse in image paragraphs. We train a multi-modal transformer and compare 11 variations of decoding algorithms. We propose to evaluate image paragraphs not only with standard automatic metrics, but also with a more extensive, "under the hood" analysis of the discourse formed by sentences. Our results show that while decoding algorithms can be unfaithful to the reference texts, they still generate grounded descriptions, but they also lack understanding of the discourse structure and differ from humans in terms of attentional structure over images. | On Decoding and Discourse Structure in Multi-Modal Text Generation |
d252624682 | This paper explores the application of the notion of transparency to annotation schemes, understood as the properties that make it easy for potential users to see the scope of the scheme, the main concepts used in annotations, and the ways these concepts are interrelated. Based on an analysis of annotation schemes in the ISO Semantic Annotation Framework, it is argued that the way these schemes make use of 'metamodels' is not optimal, since these models are often not entirely clear and not directly related to the formal specification of the scheme. It is shown that by formalizing the relation between metamodels and annotations, both can benefit and can be made simpler, and the annotation scheme becomes intuitively more transparent. | Intuitive and Formal Transparency in Semantic Annotation Schemes |
d248496316 | Self-learning paradigms in large-scale conversational AI agents tend to leverage user feedback in bridging between what they say and what they mean. However, such learning, particularly in Markov-based query rewriting systems have far from addressed the impact of these models on future training where successive feedback is inevitably contingent on the rewrite itself, especially in a continually updating environment. In this paper, we explore the consequences of this inherent lack of selfawareness towards impairing the model performance, ultimately resulting in both Type I and II errors over time. To that end, we propose augmenting the Markov Graph construction with a superposition-based adjacency matrix. Here, our method leverages an induced stochasticity to reactively learn a locally-adaptive decision boundary based on the performance of the individual rewrites in a bi-variate beta setting. We also surface a data augmentation strategy that leverages template-based generation in abridging complex conversation hierarchies of dialogs so as to simplify the learning process. All in all, we demonstrate that our selfaware model improves the overall PR-AUC by 27.45%, achieves a relative defect reduction of up to 31.22%, and is able to adapt quicker to changes in global preferences across a large number of customers. systems. | Self-Aware Feedback-Based Self-Learning in Large-Scale Conversational AI |
d5113338 | In this study we address the problem of automated word stress detection in Russian using character level models and no partspeech-taggers. We use a simple bidirectional RNN with LSTM nodes and achieve the accuracy of 90% or higher. We experiment with two training datasets and show that using the data from an annotated corpus is much more efficient than using a dictionary, since it allows us to take into account word frequencies and the morphological context of the word. | Automated Word Stress Detection in Russian |
d218487587 | This paper investigates contextual word representation models from the lens of similarity analysis. Given a collection of trained models, we measure the similarity of their internal representations and attention. Critically, these models come from vastly different architectures. We use existing and novel similarity measures that aim to gauge the level of localization of information in the deep models, and facilitate the investigation of which design factors affect model similarity, without requiring any external linguistic annotation. The analysis reveals that models within the same family are more similar to one another, as may be expected. Surprisingly, different architectures have rather similar representations, but different individual neurons. We also observed differences in information localization in lower and higher layers and found that higher layers are more affected by fine-tuning on downstream tasks. 1 | Similarity Analysis of Contextual Word Representation Models |
d252819136 | Sentiment classification is a fundamental NLP task of detecting the sentiment polarity of a given text. In this paper we show how solving sentiment span extraction as an auxiliary task can help improve final sentiment classification performance in a low-resource code-mixed setup. To be precise, we don't solve a simple multi-task learning objective, but rather design a unified transformer framework that exploits the bidirectional connection between the two tasks simultaneously. To facilitate research in this direction we release gold-standard humanannotated sentiment span extraction dataset for Tamil-english code-switched texts. Extensive experiments and strong baselines show that our proposed approach outperforms sentiment and span prediction by 1.27% and 2.78% respectively when compared to the best performing MTL baseline. We also establish the generalizability of our approach on the Twitter Sentiment Extraction dataset. We make our code and data publicly available on GitHub 1 . | Span Extraction Aided Improved Code-mixed Sentiment Classification |
d226246215 | Language pairs with limited amounts of parallel data, also known as low-resource languages, remain a challenge for neural machine translation. While the Transformer model has achieved significant improvements for many language pairs and has become the de facto mainstream architecture, its capability under low-resource conditions has not been fully investigated yet. Our experiments on different subsets of the IWSLT14 training data show that the effectiveness of Transformer under low-resource conditions is highly dependent on the hyper-parameter settings. Our experiments show that using an optimized Transformer for low-resource conditions improves the translation quality up to 7.3 BLEU points compared to using the Transformer default settings.This work is licensed under a Creative Commons Attribution 4.0 International Licence. Licence details: http:// creativecommons.org/licenses/by/4.0/. | Optimizing Transformer for Low-Resource Neural Machine Translation |
d17274515 | We define a new feature selection score for text classification based on the KL-divergence between the distribution of words in training documents and their classes. The score favors words that have a similar distribution in documents of the same class but different distributions in documents of different classes. Experiments on two standard data sets indicate that the new method outperforms mutual information, especially for smaller categories. | A New Feature Selection Score for Multinomial Naive Bayes Text Classification Based on KL-Divergence |
d202742636 | The induced action alternation and the caused-motion construction are two phenomena that allow English verbs to be interpreted as motion-causing events. This is possible when a verb is used with a direct object and a directional phrase, even when the verb does not lexically signify causativity or motion, as in "Sylvia laughed Mary off the stage". While participation in the induced action alternation is a lexical property of certain verbs, the caused-motion construction is not anchored in the lexicon. We model both phenomena with XMG-2 and use the TuLiPA parser to create compositional semantic frames for example sentences. We show how such frames represent the key differences between these two phenomena at the syntax-semantics interface, and how TAG can be used to derive distinct analyses for them. | Modeling the Induced Action Alternation and the Caused-Motion Construction with Tree Adjoining Grammar (TAG) and Semantic Frames |
d17497151 | For AI systems to reason about real world situations, they need to recognize which processes are at play and which entities play key roles in them. Our goal is to extract this kind of rolebased knowledge about processes, from multiple sentence-level descriptions. This knowledge is hard to acquire; while semantic role labeling (SRL) systems can extract sentence level role information about individual mentions of a process, their results are often noisy and they do not attempt create a globally consistent characterization of a process.To overcome this, we extend standard within sentence joint inference to inference across multiple sentences. This cross sentence inference promotes role assignments that are compatible across different descriptions of the same process. When formulated as an Integer Linear Program, this leads to improvements over within-sentence inference by nearly 3% in F1. The resulting role-based knowledge is of high quality (with a F1 of nearly 82). | Cross-Sentence Inference for Process Knowledge |
d257366180 | The task of speaker diarization aims to determine which speakers spoke when in a recording. Such functionality could help to accelerate work in endangered languages by facilitating transcription and semi-automatically extracting useful meta-data to enrich language archives. However, there has been little work on speaker diarization for low-resource or endangered languages. This work explores three neural approaches to speaker diarization applied to data sets drawn from endangered language archives. We find consistent improvements for recent neural x-vector models over earlier approaches. We also assess the factors which impact performance across models and data sets, with a focus on the challenging characteristics of endangered language recordings. | Investigating Speaker Diarization of Endangered Language Data |
d218613822 | Intelligent features in email service applications aim to increase productivity by helping people organize their folders, compose their emails and respond to pending tasks. In this work, we explore a new application, Smart-To-Do, that helps users with task management over emails. We introduce a new task and dataset for automatically generating To-Do items from emails where the sender has promised to perform an action. We design a two-stage process leveraging recent advances in neural text generation and sequenceto-sequence learning, obtaining BLEU and ROUGE scores of 0.23 and 0.63 for this task. To the best of our knowledge, this is the first work to address the problem of composing To-Do items from emails. | Smart To-Do: Automatic Generation of To-Do Items from Emails |
d186206216 | This paper presents a new approach to generating English text from Abstract Meaning Representation (AMR). In contrast to the neural and statistical MT approaches used in other AMR generation systems, this one is largely rule-based, supplemented only by a language model and simple statistical linearization models, allowing for more control over the output. We also address the difficulties of automatically evaluating AMR generation systems and the problems with BLEU for this task. We compare automatic metrics to human evaluations and show that while METEOR and TER arguably reflect human judgments better than BLEU, further research into suitable evaluation metrics is needed. | A Partially Rule-Based Approach to AMR Generation |
d51990297 | Neural machine translation has achieved impressive results in the last few years, but its success has been limited to settings with large amounts of parallel data. One way to improve NMT for lower-resource settings is to initialize a word-based NMT model with pretrained word embeddings. However, rare words still suffer from lower quality word embeddings when trained with standard word-level objectives. We introduce word embeddings that utilize morphological resources, and compare to purely unsupervised alternatives. We work with Arabic, a morphologically rich language with available linguistic resources, and perform Ar-to-En MT experiments on a small corpus of TED subtitles. We find that word embeddings utilizing subword information consistently outperform standard word embeddings on a word similarity task and as initialization of the source word embeddings in a low-resource NMT system. | Morphological Word Embeddings for Arabic Neural Machine Translation in Low-Resource Settings |
d21698065 | Paraphrases play an important role in natural language understanding, especially because there are fluent jumps between hidden paraphrases in a text. For example, even to get to the meaning of a simple dialog like I bought a computer. How much did the computer cost? involves quite a few steps. A computational system may actually have a huge problem in linking the two sentences as their connection is not overtly present in the text. However, it becomes easier if it has access to the following paraphrases:[HUMAN] buy [ARTIFACT] ⇐⇒ [HUMAN] pay [PRICE] for [ARTIFACT] ⇐⇒ [ARTIFACT] cost [HUMAN] [PRICE], and also to the information that I IsA [HUMAN] and computer IsA [ARTIFACT]. In this paper we introduce a resource of such paraphrases that was extracted by investigating large corpora in an unsupervised manner. The resource contains tens of thousands of such pairs and it is available for academic purposes. | A Large Resource for Patterns of Verbal Paraphrases |
d11775832 | Current lexical semantic theories provide representations at a coarse grained level. In this paper, I will provide motivations for a fine grained representation for verbs and.nouns. An initial case study is done to serve as evidence that a more detailed representation is needed for tasks that require high accuracy rates, such as machine translation. An automatic approach to gather fine grained information via corpus extraction is described. Lastly, issues of lexical representation and cross-lingual translation are discussed. | Inferring Semantics from Collocation Clusters to Represent Verbs and Nouns |
d1931472 | We apply cross-lingual Latent Semantic Indexing to the Bilingual Document Alignment Task at WMT16. Reduced-rank singular value decomposition of a bilingual term-document matrix derived from known English/French page pairs in the training data allows us to map monolingual documents into a joint semantic space. Two variants of cosine similarity between the vectors that place each document into the joint semantic space are combined with a measure of string similarity between corresponding URLs to produce 1:1 alignments of English/French web pages in a variety of domains. The system achieves a recall of ca. 88% if no in-domain data is used for building the latent semantic model, and 93% if such data is included.Analysing the system's errors on the training data, we argue that evaluating aligner performance based on exact URL matches under-estimates their true performance and propose an alternative that is able to account for duplicates and near-duplicates in the underlying data. | Shared Task Papers |
d252847518 | Social media analytics are widely being explored by researchers for various applications. | MUCIC at ComMA@ICON: Multilingual Gender Biased and Communal Language Identification using n-grams and Multilingual Sentence Encoders |
d12995507 | Discovering significant types of relations from the web is challenging because of its open nature. Unsupervised algorithms are developed to extract relations from a corpus without knowing the relations in advance, but most of them rely on tagging arguments of predefined types. Recently, a new algorithm was proposed to jointly extract relations and their argument semantic classes, taking a set of relation instances extracted by an open IE algorithm as input. However, it cannot handle polysemy of relation phrases and fails to group many similar ("synonymous") relation instances because of the sparseness of features. In this paper, we present a novel unsupervised algorithm that provides a more general treatment of the polysemy and synonymy problems. The algorithm incorporates various knowledge sources which we will show to be very effective for unsupervised extraction. Moreover, it explicitly disambiguates polysemous relation phrases and groups synonymous ones. While maintaining approximately the same precision, the algorithm achieves significant improvement on recall compared to the previous method. It is also very efficient. Experiments on a realworld dataset show that it can handle 14.7 million relation instances and extract a very large set of relations from the web. | Ensemble Semantics for Large-scale Unsupervised Relation Extraction |
d1935605 | Walker (1996)presents a cache model of the operation of attention in the processing of discourse as an alternative to the focus space stack that was proposed previously by Grosz and Sidner (Grosz 1977a;Grosz and Sidner 1986). In this squib, we present a critical analysis of the cache model and of Walker's supporting evidence from anaphora in discourses with interruptions and from informationally redundant utterances. We argue that the cache model is underdetermined in several ways that are crucial to a comparison of the two models and conclude that Walker has not established the superiority of the cache model. We also argue that psycholinguistic evidence does not support the cache model over the focus stack model.Attentional State, Focus, and Limited CapacityWalker argues that "[t]he notion of a cache in combination with main memory, as is standard in computational architectures, is a good basis for a computational model of human attentional capacity in discourse processing (emphasis added)" (page 258). Walker further claims that this model has advantages in explaining discourse phenomena over what she refers to as the "stack model." * Preparation of this paper was supported by NSF grants IRI-94-04756 and IIS-9811129. | Conceptions of Limited Attention and Discourse Focus |
d218974139 | ||
d6614720 | Lexical semantic resources, like WordNet, are often used in real applications of natural language document processing. For example, we integrated GermaNet in our document suite XDOC. In addition to hypernymy and synonymy relations, we want to exploit GermaNet verb frames for our analysis. In this paper, we outline an approach for the domain related enrichment of GermaNet verb frames by corpus based syntactic and co-occurrence data analyses of real documents. | Corpus based Enrichment of GermaNet Verb Frames |
d8269809 | The paper proposed a new syntactic annotation scheme ---functional chunk, which tried to represent information about grammatical relations between sentence-level predicates and their arguments. Under this scheme, we built a Chinese chunk bank with about two million Chinese characters, and developed some learned models for automatically annotating fresh text with functional chunks. We also proposed a two-stages approach to build Chinese tree bank on the top of chunk bank, and gave some experimental results of chunk-based syntactic parser to show the advantage of functional chunk for parsing performance increase. All these work lays good foundations for further research project to build a large scale Chinese tree bank. | Annotating the functional chunks in Chinese sentences |
d60689 | We describe the University of Illinois (UI) system that participated in the Helping Our Own (HOO) 2012 shared task, which focuses on correcting preposition and determiner errors made by non-native English speakers. The task consisted of three metrics: Detection, Recognition, and Correction, and measured performance before and after additional revisions to the test data were made. Out of 14 teams that participated, our system scored first in Detection and Recognition and second in Correction before the revisions; and first in Detection and second in the other metrics after revisions. We describe our underlying approach, which relates to our previous work in this area, and propose an improvement to the earlier method, error inflation, which results in significant gains in performance. | The 7th Workshop on the Innovative Use of NLP for Building Educational Applications, pages 272-280, The UI System in the HOO 2012 Shared Task on Error Correction |
d252819315 | Bridging reference resolution is the task of finding nouns that complement essential information of another noun. The essentiality varies depending on noun combination and context and has a continuous distribution. Despite the continuous nature of essentiality, existing datasets of bridging reference have only a few coarse labels to represent the essentiality(Poesio and Artstein, 2008;Hangyo et al., 2012). In this work, we propose a crowdsourcing-based annotation method that considers continuous essentiality. In the crowdsourcing task, we asked workers to select both all nouns with a bridging reference relation and a noun with the highest essentiality among them. Combining these annotations, we can obtain continuous essentiality. Experimental results demonstrated that the constructed dataset improves bridging reference resolution performance. The code is available at https://github.com/nobu-g/ bridging-resolution. | Improving Bridging Reference Resolution using Continuous Essentiality from Crowdsourcing |
d12123792 | In this paper we present a gold standard dataset for Entity Linking (EL) in the Music Domain. It contains thousands of musical named entities such as Artist, Song or Record Label, which have been automatically annotated on a set of artist biographies coming from the Music website and social network LAST.FM. The annotation process relies on the analysis of the hyperlinks present in the source texts and in a voting-based algorithm for EL, which considers, for each entity mention in text, the degree of agreement across three state-of-the-art EL systems. Manual evaluation shows that EL Precision is at least 94%, and due to its tunable nature, it is possible to derive annotations favouring higher Precision or Recall, at will. We make available the annotated dataset along with evaluation data and the code. | ELMD: An Automatically Generated Entity Linking Gold Standard Dataset in the Music Domain |
d250390936 | The performance of Reinforcement Learning (RL) for natural language tasks including Machine Translation (MT) is crucially dependent on the reward formulation. This is due to the intrinsic difficulty of the task in the high-dimensional discrete action space as well as the sparseness of the standard reward functions defined for limited set of groundtruth sequences biased towards singular lexical choices. To address this issue, we formulate SURF, a maximally dense semantic-level unsupervised reward function which mimics human evaluation by considering both sentence fluency and semantic similarity. We demonstrate the strong potential of SURF to leverage a family of Actor-Critic Transformerbased Architectures with synchronous and asynchronous multi-agent variants. To tackle the problem of large action-state spaces, each agent is equipped with unique exploration strategies, promoting diversity during its exploration of the hypothesis space. When BLEU scores are compared, our dense unsupervised reward outperforms the standard sparse reward by 2% on average for in-and out-of-domain settings. | SURF: Semantic-level Unsupervised Reward Function for Machine Translation |
d17683994 | Community question answering (CQA) systems such as Yahoo! Answers allow registered-usersto ask and answer questions in various question categories. However, a significant percentage of asked questions in Yahoo! Answers are unanswered. In this paper, we propose to reduce this percentage by reusing answers to past resolved questions from the site. Specifically, we propose to satisfy unanswered questions in entity rich categories by searching for and reusing the best answers to past resolved questions with shared needs. For unanswered questions that do not have a past resolved question with a shared need, we propose to use the best answer to a past resolved question with similar needs. Our experiments on a Yahoo! Answers dataset shows that our approach retrieves most of the past resolved questions that have shared or similar needs to unanswered questions. | An Entity-Based approach to Answering Recurrent and Non-Recurrent Questions with Past Answers |
d226283455 | The recent paradigm shift to contextual word embeddings has seen tremendous success across a wide range of down-stream tasks. However, little is known on how the emergent relation of context and semantics manifests geometrically. We investigate polysemous words as one particularly prominent instance of semantic organization.Our rigorous quantitative analysis of linear separability and cluster organization in embedding vectors produced by BERT shows that semantics do not surface as isolated clusters but form seamless structures, tightly coupled with sentiment and syntax. | How does BERT capture semantics? A closer look at polysemous words |
d15644904 | We briefly describe a two-way speech-tospeech English-Farsi translation system prototype developed for use in doctorpatient interactions. The overarching philosophy of the developers has been to create a system that enables effective communication, rather than focusing on maximizing component-level performance. The discussion focuses on the general approach and evaluation of the system by an independent government evaluation team. | Transonics: A Practical Speech-to-Speech Translator for English-Farsi Medical Dialogues |
d2814545 | The problem of finding the most probable string for a distribution generated by a weighted finite automaton or a probabilistic grammar is related to a number of important questions: computing the distance between two distributions or finding the best translation (the most probable one) given a probabilistic finite state transducer. The problem is undecidable with general weights and is N P-hard if the automaton is probabilistic. We give a pseudo-polynomial algorithm which computes the most probable string in time polynomial in the inverse of the probability of the most probable string itself, both for probabilistic finite automata and probabilistic context-free grammars. We also give a randomised algorithm solving the same problem. | Finding the Most Probable String and the Consensus String: an Algorithmic Study |
d210936463 | Natural language inference (NLI) is a key part of natural language understanding. The NLI task is defined as a decision problem whether a given sentence -hypothesis -can be inferred from a given text. Typically, we deal with a text consisting of just a single premise/single sentence, which is called a single premise entailment (SPE) task. Recently, a derived task of NLI from multiple premises (MPE) was introduced together with the first annotated corpus and corresponding several strong baselines. Nevertheless, the further development in MPE field requires accessibility of huge amounts of annotated data. In this paper we introduce a novel method for rapid deriving of MPE corpora from an existing NLI (SPE) annotated data that does not require any additional annotation work. This proposed approach is based on using an open information extraction system. We demonstrate the application of the method on a well known SNLI corpus. Over the obtained corpus, we provide the first evaluations as well as we state a strong baseline. | Exploiting Open IE for Deriving Multiple Premises Entailment Corpus |
d204783696 | In this paper we describe the early stage application of the Universal Dependencies to an Italian corpus from social media developed for shared tasks related to irony and stance detection. The development of this novel resource (TWITTIRÒ-UD) serves a twofold goal: it enriches the scenario of treebanks for social media and for Italian, and it paves the way for a more reliable extraction of a larger variety of morphological and syntactic features to be used by sentiment analysis tools. On the one hand, social media texts are especially hard to parse and the limited amount of resources for training and testing NLP tools further damages the situation. On the other hand, we thought that adding the Universal Dependencies format to the fine-grained annotation for irony, that was previously applied on TWITTIRÒ, might meaningfully help in the investigation of possible relationships between syntax and semantics of the uses of figurative language, irony in particular. | Presenting TWITTIRÒ-UD: An Italian Twitter Treebank in Universal Dependencies |
d10392751 | The system to spot INIs, DNIs and their antecedents is an adaptation of VENSES, a system for semantic evaluation that has been used for RTE challenges in the last 6 years. In the following we will briefly describe the system and then the additions we made to cope with the new task. In particular, we will discuss how we mapped the VENSES analysis to the representation of frame information in order to identify null instantiations in the text. | VENSES++: Adapting a deep semantic processing system to the identification of null instantiations |
d15182936 | We report on first annotation experiments on narrative segments. Narrative segments are a pragmatic intermediate layer that allows studying more complex narratological phenomena. Our experiments show that segmenting on limited context information alone is difficult. High interannotator agreement on this task can be achieved by coupling the segmentation with summarization and aligning parts of the summaries to segments of the text. | Towards Annotating Narrative Segments |
d1090122 | We describe an attention-based convolutional neural network for the English semantic textual similarity (STS) task in the SemEval-2016 competition (Agirre et al., 2016). We develop an attention-based input interaction layer and incorporate it into our multiperspective convolutional neural network(He et al., 2015), using the PARAGRAM-PHRASE word embeddings(Wieting et al., 2016)trained on paraphrase pairs. Without using any sparse features, our final model outperforms the winning entry in STS2015 when evaluated on the STS2015 data. | UMD-TTIC-UW at SemEval-2016 Task 1: Attention-Based Multi-Perspective Convolutional Neural Networks for Textual Similarity Measurement |
d1552251 | This paper describes a new approach to automatic learning of strategies for social multi-user human-robot interaction. Using the example of a robot bartender that tracks multiple customers, takes their orders, and serves drinks, we propose a model consisting of a Social State Recogniser (SSR) which processes audio-visual input and maintains a model of the social state, together with a Social Skills Executor (SSE) which takes social state updates from the SSR as input and generates robot responses as output. The SSE is modelled as two connected Markov Decision Processes (MDPs) with action selection policies that are jointly optimised in interaction with a Multi-User Simulation Environment (MUSE). The SSR and SSE have been integrated in the robot bartender system and evaluated with human users in hand-coded and trained SSE policy variants. The results indicate that the trained policy outperformed the hand-coded policy in terms of both subjective (+18%) and objective (+10.5%) task success. | Training and evaluation of an MDP model for social multi-user human-robot interaction |
d247626326 | When reading stories, people can naturally identify sentences in which a new event starts, i.e., event boundaries, using their knowledge of how events typically unfold, but a computational model to detect event boundaries is not yet available. We characterize and detect sentences with expected or surprising event boundaries in an annotated corpus of short diary-like stories, using a model that combines commonsense knowledge and narrative flow features with a RoBERTa classifier. Our results show that, while commonsense and narrative features can help improve performance overall, detecting event boundaries that are more subjective remains challenging for our model. We also find that sentences marking surprising event boundaries are less likely to be causally related to the preceding sentence, but are more likely to express emotional reactions of story characters, compared to sentences with no event boundary. | Uncovering Surprising Event Boundaries in Narratives |
d44252659 | ||
d14054366 | This paper discusses word choice for natural language generation. It examines 11 issues, the solutions that have been proposed for them, and their implications for design. The issues are: | Issues in Word Choice |
d252125483 | Automatic morphology induction is important for computational processing of natural language. In resource-scarce languages in particular, it offers the possibility of supplementing data-driven strategies of Natural Language Processing with morphological rules that may cater for out-of-vocabulary words. Unfortunately, popular approaches to unsupervised morphology induction do not work for some of the most productive morphological processes of the Yorùbá language. To the best of our knowledge, the automatic induction of such morphological processes as full and partial reduplication, infixation, interfixation, compounding and other morphological processes, particularly those based on the affixation of stem-derived morphemes have not been adequately addressed in the literature. This study proposes a method for the automatic detection of stem-derived morphemes in Yorùbá. Words in a Yorùbá lexicon of 14,670 word-tokens were clustered around "word-labels". A word-label is a textual proxy of the patterns imposed on words by the morphological processes through which they were formed. Results confirm a conjectured significant difference between the predicted and observed probabilities of word-labels motivated by stem-derived morphemes. This difference was used as basis for automatic identification of words formed by the affixation of stem-derived morphemes. | Automatic Detection of Morphological Processes in the Yorùbá Language |
d14983699 | As the amount of cultural data available on the Semantic Web is expanding, the demand of accessing this data in multiple languages is increasing. Previous work on multilingual access to cultural heritage information has shown that at least two different problems must be dealt with when mapping from ontologies to natural language: (1) mapping multilingual metadata to interoperable knowledge sources; (2) assigning multilingual knowledge to cultural data. This paper presents our effort to deal with these problems. We describe our experiences with processing museum data extracted from two distinct sources, harmonizing this data and making its content accessible in natural language. We extend prior work in two ways. First, we present a grammar-based system that is designed to generate coherent texts from Semantic Web ontologies in 15 languages. Second, we describe how this multilingual system is exploited to form queries using the standard query language SPARQL. The generation and retrieval system builds on W3C standards and is available for further research. | Multilingual access to cultural heritage content on the Semantic Web |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.