_id stringlengths 4 10 | text stringlengths 0 18.4k | title stringlengths 0 8.56k |
|---|---|---|
d219310223 | ||
d18589806 | With the advent of word representations, word similarity tasks are becoming increasing popular as an evaluation metric for the quality of the representations. In this paper, we present manually annotated monolingual word similarity datasets of six Indian languages -Urdu, Telugu, Marathi, Punjabi, Tamil and Gujarati. These languages are most spoken Indian languages worldwide after Hindi and Bengali. For the construction of these datasets, our approach relies on translation and re-annotation of word similarity datasets of English. We also present baseline scores for word representation models using state-of-the-art techniques for Urdu, Telugu and Marathi by evaluating them on newly created word similarity datasets. | Word Similarity Datasets for Indian Languages: Annotation and Baseline Systems |
d220058074 | ||
d199368325 | ||
d18046005 | This paper presents a new approach to syntactic disambiguation based on lexicalized grammars. While existing disambiguation models decompose the probability of parsing results into that of primitive dependencies of two words, our model selects the most probable parsing result from a set of candidates allowed by a lexicalized grammar. Since parsing results given by the lexicalized grammar cannot be decomposed into independent sub-events, we apply a maximum entropy model for feature forests, which allows probabilistic modeling without the independence assumption. Our approach provides a general method of producing a consistent probabilistic model of parsing results given by lexicalized grammars. | A model of syntactic disambiguation based on lexicalized grammars |
d5569514 | Range Concatenation Grammars (RCGs) are a syntactic formalism which possesses many attractive properties. It is more powerful than Linear Context-Free Rewriting Systems, though this power is not reached to the detriment of efficiency since its sentences can always be parsed in polynomial time. If the input, instead of a string, is a Directed Acyclic Graph (DAG), only simple RCGs can still be parsed in polynomial time. For non-linear RCGs, this polynomial parsing time cannot be guaranteed anymore. In this paper, we show how the standard parsing algorithm can be adapted for parsing DAGs with RCGs, both in the linear (simple) and in the non-linear case. | Parsing Directed Acyclic Graphs with Range Concatenation Grammars |
d52289318 | This paper proposes a modularized sense induction and representation learning model that jointly learns bilingual sense embeddings that align well in the vector space, where the crosslingual signal in the English-Chinese parallel corpus is exploited to capture the collocation and distributed characteristics in the language pair. The model is evaluated on the Stanford Contextual Word Similarity (SCWS) dataset to ensure the quality of monolingual sense embeddings. In addition, we introduce Bilingual Contextual Word Similarity (BCWS), a large and high-quality dataset for evaluating crosslingual sense embeddings, which is the first attempt of measuring whether the learned embeddings are indeed aligned well in the vector space. The proposed approach shows the superior quality of sense embeddings evaluated in both monolingual and bilingual spaces. 1 | CLUSE: Cross-Lingual Unsupervised Sense Embeddings |
d9325119 | We apply statistical methods to perform automatic extraction of Hungarian collocations from corpora. Due to the complexity of Hungarian morphology, a complex resource preparation tool chain has been developed. This tool chain implements a reusable and, in principle, language independent framework. In the first part, the paper describes the tool chain itself, then, in the second part, an experiment using this framework. The experiment deals with the extraction of <verb+noun+casemark> patterns from the corpus as collocation candidates, in order to compare results to an experiment on Dutch V + PP patterns(Villada, 2004). Statistical processing on this dataset provided interesting observations, briefly explained in the evaluation section. We conclude by providing a summary of further steps required to improve the extraction process. This is not restricted to improvements in the resource preparation for statistical processing, but a proposal to use nonstatistical means as well, thus acquiring an efficient blend of different methods. | A New Approach to the Corpus-based Statistical Investigation of Hungarian Multi-word Lexemes |
d236144991 | ||
d258170298 | We present the first Africentric SemEval Shared task, Sentiment Analysis for African Languages (AfriSenti-SemEval) 1 . AfriSenti-SemEval is a sentiment classification challenge in 14 African languages (Amharic, Algerian Arabic, Hausa, Igbo, Kinyarwanda, Moroccan Arabic, Mozambican Portuguese, Nigerian Pidgin, Oromo, Swahili, Tigrinya, Twi, Xitsonga, and Yorùbá)(Muhammad et al., 2023), using data labeled with 3 sentiment classes. We present three subtasks: (1) Task A: monolingual classification, which received 44 submissions; (2) Task B: multilingual classification, which received 32 submissions; and (3) Task C: zero-shot classification, which received 34 submissions. The best performance for tasks A and B was achieved by NLNDE team with 71.31 and 75.06 weighted F1, respectively. UCAS-IIE-NLP achieved the best average score for task C with 58.15 weighted F1. We describe the various approaches adopted by the top 10 systems and their approaches. | SemEval-2023 Task 12: Sentiment Analysis for African Languages (AfriSenti-SemEval) |
d12373614 | In this paper we present an approach to tuning of context features acquired from corpora. The approach is based on the idea of a genetic algorithm (GA). We analyse a whole population of contexts surrounding related linguistic entities in order to find a generic property characteristic of such contexts. Our goal is to tune the context properties so as not to lose any correct feature values, but also to minimise the presence of ambiguous values. The GA implements a crossover operator based on dominant and recessive genes, where a gene corresponds to a context feature. A dominant gene is the one that, when combined with another gene of the same type, is inevitably reflected in the offspring. Dominant genes denote the more suitable context features. In each iteration of the GA, the number of individuals in the population is halved, finally resulting in a single individual that contains context features tuned with respect to the information contained in the training corpus. We illustrate the general method by using a case study concerned with the identification of relationships between verbs and terms complementing them. More precisely, we tune the classes of terms that are typically selected as arguments for the considered verbs in order to acquire their semantic features. | Tuning Context Features with Genetic Algorithms * |
d14002118 | In this paper we describe the mapping ofZaliznjak's (1977)morphological classes into the lexical representation language DATR (Evans and Gazdar 1996). On the basis of the resulting DATR theory a set of fully inflected forms together with their associated morphosyntax can automatically be generated from the electronic version of Zaliznjak's dictionary(Ilola and Mustajoki 1989). From this data we plan to develop a wide-coverage morphosyntactic lemmatizer and tagger for Russian. | A large-scale inheritance-based morphological lexicon for Russian |
d10178418 | In this paper we report on a set of computational tools with (n)SGML pipeline data flow for uncovering internal structure in natural language texts. The main idea behind the workbench is the independence of the text representation and text analysis phases. At the representation phase the text is converted from a sequence of characters to features of interest by means of the annotation tools. At the analysis phase those features are used by statistics gathering and inference tools for finding significant correlations in the texts. The analysis tools are independent of particular assumptions about the nature of the feature-set and work on the abstract level of featureelements represented as SGML items. | A Workbench for Finding Structure in Texts |
d6813036 | This paper describes a method for obtaining the semantic representation for a syntax tree in Systemic Grammar (SG). A prototype implementation of this method --the REVELATION1 semantic interpreter --has been ldeveloped. It is derived from a SG generator foe a large subset of English --GENESYS --and is !thus, in contrast with most reversible grammars, fan interpreter based on a generator. A task decomposition approach is adopted for this reversal process which operates within the framework of SG, thus demonstrating that Systemic Grammars can be reversed and hence that a SG is a truly bi-directional formalism, p | A :SEMANTIC INTERPRETER FOR SYSTEMIC GRAMMARS |
d39661249 | How few is too few? Determining the minimum acceptable number of LSA dimensions to visualise text cohesion with Lex Abstract | |
d2281198 | Most research on automated categorization of documents has concentrated on the assignment of one or many categories to a whole text. However, new applications, e.g. in the area of the Semantic Web, require a richer and more fine-grained annotation of documents, such as detailed thematic information about the parts of a document. Hence we investigate the automatic categorization of text segments of scientific articles with XML markup into 16 topic types from a text type structure schema. A corpus of 47 linguistic articles was provided with XML markup on different annotation layers representing text type structure, logical document structure, and grammatical categories. Six different feature extraction strategies were applied to this corpus and combined in various parametrizations in different classifiers. The aim was to explore the contribution of each type of information, in particular the logical structure features, to the classification accuracy. The results suggest that some of the topic types of our hierarchy are successfully learnable, while the features from the logical structure layer had no particular impact on the results. | Text Type Structure and Logical Document Structure |
d253761257 | The Event Causality Identification Shared Task of CASE 2022 involved two subtasks working on the Causal News Corpus. Subtask 1 required participants to predict if a sentence contains a causal relation or not. This is a supervised binary classification task. Subtask 2 required participants to identify the Cause, Effect and Signal spans per causal sentence. This could be seen as a supervised sequence labeling task. For both subtasks, participants uploaded their predictions for a held-out test set, and ranking was done based on binary F1 and macro F1 scores for Subtask 1 and 2, respectively. This paper summarizes the work of the 17 teams that submitted their results to our competition and 12 system description papers that were received. The best F1 scores achieved for Subtask 1 and 2 were 86.19% and 54.15%, respectively. All the top-performing approaches involved pretrained language models fine-tuned to the targeted task. We further discuss these approaches and analyze errors across participants' systems in this paper. | Event Causality Identification with Causal News Corpus -Shared Task 3, CASE 2022 |
d259202515 | Audio-visual speech recognition (AVSR) provides a promising solution to ameliorate the noise-robustness of audio-only speech recognition with visual information. However, most existing efforts still focus on audio modality to improve robustness considering its dominance in AVSR task, with noise adaptation techniques such as front-end denoise processing. Though effective, these methods are usually faced with two practical challenges: 1) lack of sufficient labeled noisy audio-visual training data in some real-world scenarios and 2) less optimal model generality to unseen testing noises. In this work, we investigate the noiseinvariant visual modality to strengthen robustness of AVSR, which can adapt to any testing noises while without dependence on noisy training data, a.k.a., unsupervised noise adaptation. Inspired by human perception mechanism, we propose a universal viseme-phoneme mapping (UniVPM) approach to implement modality transfer, which can restore clean audio from visual signals to enable speech recognition under any noisy conditions. Extensive experiments on public benchmarks LRS3 and LRS2 show that our approach achieves the state-of-the-art under various noisy as well as clean conditions. In addition, we also outperform previous stateof-the-arts on visual speech recognition task 1 . | Hearing Lips in Noise: Universal Viseme-Phoneme Mapping and Transfer for Robust Audio-Visual Speech Recognition |
d247619041 | Few-shot NER needs to effectively capture information from limited instances and transfer useful knowledge from external resources. In this paper, we propose a self-describing mechanism for few-shot NER, which can effectively leverage illustrative instances and precisely transfer knowledge from external resources by describing both entity types and mentions using a universal concept set. Specifically, we design Self-describing Networks (SDNet), a Seq2Seq generation model which can universally describe mentions using concepts, automatically map novel entity types to concepts, and adaptively recognize entities ondemand. We pre-train SDNet with large-scale corpus, and conduct experiments on 8 benchmarks from different domains. Experiments show that SDNet achieves competitive performances on all benchmarks and achieves the new state-of-the-art on 6 benchmarks, which demonstrates its effectiveness and robustness. | Few-shot Named Entity Recognition with Self-describing Networks |
d258309749 | Despite significant progress in Natural Language Generation for Indian languages (Indic-NLP), there is a lack of datasets around complex structured tasks such as semantic parsing. One reason for this imminent gap is the complexity of the logical form, which makes English to multilingual translation difficult. The process involves alignment of logical forms, intents and slots with translated unstructured utterance. To address this, we propose an Interbilingual Seq2seq Semantic parsing dataset IE-SEMPARSE for 11 distinct Indian languages. We highlight the proposed task's practicality, and evaluate existing multilingual seq2seq models across several train-test strategies. Our experiment reveals a high correlation across performance of original multilingual semantic parsing datasets (such as mTOP, multilingual TOP and multiATIS++) and our proposed IE-SEMPARSE suite. | Evaluating Inter-Bilingual Semantic Parsing for Indian Languages |
d2296109 | The alignment of syntactic trees is the task of aligning the internal and leaf nodes of two sentences in different languages structured as trees. The output of the alignment can be used, for instance, as knowledge resource for learning translation rules (for rule-based machine translation systems) or models (for statistical machine translation systems). This paper presents some experiments carried out based on two syntactic tree alignment algorithms presented in[Lavie et al. 2008] and[Tinsley et al. 2007]. Aiming at improving the performance of internal nodes alignment, some approaches for combining the output of these two algorithms were evaluated in Brazilian Portuguese and English parallel trees. | Combining Models for the Alignment of Parallel Syntactic Trees |
d7559647 | In this demonstration we present WoSIT, an API for Word Sense Induction (WSI) algorithms. The toolkit provides implementations of existing graph-based WSI algorithms, but can also be extended with new algorithms. The main mission of WoSIT is to provide a framework for the extrinsic evaluation of WSI algorithms, also within end-user applications such as Web search result clustering and diversification. | WoSIT: A Word Sense Induction Toolkit for Search Result Clustering and Diversification |
d259137386 | Context information modeling is an important task in conversational KBQA. However, existing methods usually assume the independence of utterances and model them in isolation. In this paper, we propose a History Semantic Graph Enhanced KBQA model (HSGE) that is able to effectively model long-range semantic dependencies in conversation history while maintaining low computational cost. The framework incorporates a context-aware encoder, which employs a dynamic memory decay mechanism and models context at different levels of granularity. We evaluate HSGE on a widely used benchmark dataset for complex sequential question answering. Experimental results demonstrate that it outperforms existing baselines averaged on all question types. | History Semantic Graph Enhanced Conversational KBQA with Temporal Information Modeling |
d196391 | Extraction and interpretation of temporal information from clinical text is essential for clinical practitioners and researchers. SemEval 2016 Task 12 (Clinical TempEval) addressed this challenge using the THYME 1 corpus, a corpus of clinical narratives annotated with a schema based on TimeML 2 guidelines. We developed and evaluated approaches for: extraction of temporal expressions (TIMEX3) and EVENTs; TIMEX3 and EVENT attributes; document-time relations; and narrative container relations. Our approach is based on supervised learning (CRF and logistic regression), utilizing various sets of syntactic, lexical and semantic features with addition of manually crafted rules. Our system demonstrated substantial improvements over the baselines in all the tasks. | GUIR at SemEval-2016 task 12: Temporal Information Processing for Clinical Narratives |
d14580933 | Ørÿ=.]oÏ>ï áRI A Survey of Full-text Data Bases and Related Techniques for Chinese Ancient Documents | |
d1604930 | In this paper, we describe our system submitted for the semantic textual similarity (STS) task at SemEval 2012. We implemented two approaches to calculate the degree of similarity between two sentences. First approach combines corpus-based semantic relatedness measure over the whole sentence with the knowledge-based semantic similarity scores obtained for the words falling under the same syntactic roles in both the sentences. We fed all these scores as features to machine learning models to obtain a single score giving the degree of similarity of the sentences. Linear Regression and Bagging models were used for this purpose. We used Explicit Semantic Analysis (ESA) as the corpus-based semantic relatedness measure. For the knowledgebased semantic similarity between words, a modified WordNet based Lin measure was used. Second approach uses a bipartite based method over the WordNet based Lin measure, without any modification. This paper shows a significant improvement in calculating the semantic similarity between sentences by the fusion of the knowledge-based similarity measure and the corpus-based relatedness measure against corpus based measure taken alone. | DERI&UPM: Pushing Corpus Based Relatedness to Similarity: Shared Task System Description |
d1423474 | PREL1MIN A R IESWe study the formal and linguistic properties of a class of parenthesis-free categorial grammars derived from those of Ades and Steedman by varying the set of reduction rules. We characterize the reduction rules capable of generating context-sensitive languages as those having a partial combination rule and a combination rule in the reverse direction. We show that any categorial language is a permutation of some context-free language, thus inheriting properties dependent on symbol counting only. We compare some of their properties with other contemporary formalisms. | CATEGORIAL AND NON-CATEGORIAL LANGUAGES |
d235097203 | ||
d221373747 | Nous présentons le système utilisé par l'équipe Synapse/IRIT dans la compétition DEFT2019 portant sur deux tâches liées à des cas cliniques rédigés en français : l'une d'appariement entre des cas cliniques et des discussions, l'autre d'extraction de mots-clefs. Une des particularité est l'emploi d'apprentissage non-supervisé sur les deux tâches, sur un corpus construit spécifiquement pour le domaine médical en français ABSTRACT Unsupervised learning for matching and labelling of french clincal cases -DEFT2019 We present the system used by the Synapse / IRIT team in the DEFT2019 competition covering two tasks on clinical cases written in French : the matching between clinical cases and discussions, and the extraction of key words. A particularity of our submissions is the use of unsupervised learning on both tasks, thanks to a french corpus of medical texts we gathered. MOTS-CLÉS : TALN bio-médical, DEFT2019, apprentissage non-supervisé. | |
d5677636 | In many languages such as Bambara or Arabic, tone markers (diacritics) may be written but are actually often omitted. NLP applications are confronted to ambiguities and subsequent difficulties when processing texts. To circumvent this problem, tonalization may be used, as a word sense disambiguation task, relying on context to add diacritics that partially disambiguate words as well as senses. In this paper, we describe our implementation of a Bambara tonalizer that adds tone markers using machine learning (CRFs). To make our tool efficient, we used differential coding, word segmentation and edit operation filtering. We describe our approach that allows tractable machine learning and improves accuracy: our model may be learned within minutes on a 358K-word corpus and reaches 92.3% accuracy. | A Bambara Tonalization System for Word Sense Disambiguation Using Differential Coding, Segmentation and Edit Operation Filtering |
d237155023 | ||
d3878510 | This paper presents two strong baselines for the BioNLP 2009 shared task on event extraction. First we re-implement a rulebased approach which allows us to explore the task and the effect of domainadapted parsing on it. We then replace the rule-based component with support vector machine classifiers and achieve performance near the state-of-the-art without using any external resources. The good performances achieved and the relative simplicity of both approaches make them reproducible baselines. We conclude with suggestions for future work with respect to the task representation. | Two strong baselines for the BioNLP 2009 event extraction task |
d46151820 | ||
d258461017 | Question answering over knowledge bases is considered a difficult problem due to the challenge of generalizing to a wide variety of possible natural language questions. Additionally, the heterogeneity of knowledge base schema items between different knowledge bases often necessitates specialized training for different knowledge base question-answering (KBQA) datasets. To handle questions over diverse KBQA datasets with a unified training-free framework, we propose KB-BINDER, which for the first time enables few-shot in-context learning over KBQA tasks. Firstly, KB-BINDER leverages large language models like Codex to generate logical forms as the draft for a specific question by imitating a few demonstrations. Secondly, KB-BINDER grounds on the knowledge base to bind the generated draft to an executable one with BM25 score matching. The experimental results on four public heterogeneous KBQA datasets show that KB-BINDER can achieve a strong performance with only a few in-context demonstrations. Especially on GraphQA and 3-hop MetaQA, KB-BINDER can even outperform the state-of-theart trained models. On GrailQA and WebQSP, our model is also on par with other fullytrained models. We believe KB-BINDER can serve as an important baseline for future research. Our code is available at https:// github.com/ltl3A87/KB-BINDER | Few-shot In-context Learning for Knowledge Base Question Answering |
d6483979 | We present a hierarchical statistical machine translation system which supports discontinuous constituents. It is based on synchronous linear context-free rewriting systems (SLCFRS), an extension to synchronous context-free grammars in which synchronized non-terminals span k ≥ 1 continuous blocks on either side of the bitext. This extension beyond contextfreeness is motivated by certain complex alignment configurations that are beyond the alignment capacity of current translation models and their relatively frequent occurrence in hand-aligned data. Our experiments for translating from German to English demonstrate the feasibility of training and decoding with more expressive translation models such as SLCFRS and show a modest improvement over a context-free baseline. | Hierarchical Machine Translation With Discontinuous Phrases |
d11745386 | In this paper we describe the construction of a parallel corpus between the standard and a non-standard language variety, specifically standard Austrian German and Viennese dialect. The resulting parallel corpus is used for statistical machine translation (SMT) from the standard to the non-standard variety. The main challenges to our task are data scarcity and the lack of an authoritative orthography. We started with the generation of a base corpus of manually transcribed and translated data from spoken text encoded in a specifically developed orthography. This data is used to train a first phrasebased SMT. To deal with out-of-vocabulary items we exploit the strong proximity between source and target variety with a backoff strategy that uses character-level models. To arrive at the necessary size for a corpus to be used for SMT, we employ a boot-strapping approach. Integrating additional available sources (comparable corpora, such as Wikipedia) necessitates to identify parallel sentences out of substantially differing parallel documents. As an additional task, the spelling of the texts has to be transformed into the above mentioned orthography of the target variety. | Corpus development for machine translation between standard and dialectal varieties |
d260063216 | Code-mixing refers to the phenomenon of using two or more languages interchangeably within a speech or discourse context. This practice is particularly prevalent on social media platforms, and determining the embedded affects in a code-mixed sentence remains as a challenging problem. In this submission we describe our system for WASSA 2023 Shared Task on Emotion Detection in English-Urdu code-mixed text. In our system we implement a multiclass emotion detection model with label space of 11 emotions. Samples are code-mixed English-Urdu text, where Urdu is written in romanised form. Our submission is limited to one of the subtasks -Multi Class classification and we leverage transformer-based Multilingual Large Language Models (MLLMs), XLM-RoBERTa and Indic-BERT. We fine-tune MLLMs on the released data splits, with and without pre-processing steps (translation to english), for classifying texts into the appropriate emotion category. Our methods did not surpass the baseline, and our submission is ranked sixth overall. | PrecogIIITH@WASSA2023: Emotion Detection for Urdu-English Code-mixed Text |
d227905455 | ||
d253761988 | Sentence acceptability judgment assesses to what degree a sentence is acceptable to native speakers of the language. Most unsupervised prediction approaches rely on a language model to obtain the likelihood of a sentence that reflects acceptability. However, two problems exist: first, low-frequency words would have a significant negative impact on the sentence likelihood derived from the language model; second, when it comes to multiple domains, the language model needs to be trained on domain-specific text for domain adaptation. To address both problems, we propose a simple method that substitutes Part-of-Speech (POS) tags for low-frequency words in sentences used for continual training of masked language models. Experimental results show that our word-tag-hybrid BERT model brings improvement on both a sentence acceptability benchmark and a cross-domain sentence acceptability evaluation corpus. Furthermore, our annotated cross-domain sentence acceptability evaluation corpus would benefit future research. | A Simple Yet Effective Hybrid Pre-trained Language Model for Unsupervised Sentence Acceptability Prediction |
d17560522 | This paper describes the goals, design and results of a shared task on the automatic linguistic annotation of German language data from genres of computer-mediated communication (CMC), social media interactions and Web corpora. The two subtasks of tokenization and part-of-speech tagging were performed on two data sets: (i) a genuine CMC data set with samples from several CMC genres, and (ii) a Web corpora data set of CC-licensed Web pages which represents the type of data found in large corpora crawled from the Web. The teams participating in the shared task achieved a substantial improvement over current off-the-shelf tools for German. The best tokenizer reached an F 1score of 99.57% (vs. 98.95% off-the-shelf baseline), while the best tagger reached an accuracy of 90.44% (vs. 84.86% baseline). The gold standard (more than 20,000 tokens of training and test data) is freely available online together with detailed annotation guidelines. | EmpiriST 2015: A Shared Task on the Automatic Linguistic Annotation of Computer-Mediated Communication and Web Corpora |
d227231263 | ||
d13057309 | This paper describes a system that attempts to interpret descriptive texts without the uJe of complex grammars. The purpose of the system is to transform the descriptions to a standard form which may be used as the basis of a database system knowledgeable in the subject matter of the teXt. | AUTOMATIC ANALYSIS OF DESCRIPTIVE TEXTS |
d252968038 | We present Expected Statistic Regulariza tion (ESR), a novel regularization technique that utilizes low-order multi-task structural statistics to shape model distributions for semisupervised learning on low-resource datasets. We study ESR in the context of cross-lingual transfer for syntactic analysis (POS tagging and labeled dependency parsing) and present several classes of low-order statistic functions that bear on model behavior. Experimentally, we evaluate the proposed statistics with ESR for unsupervised transfer on 5 diverse target languages and show that all statistics, when estimated accurately, yield improvements to both POS and LAS, with the best statistic improving POS by +7.0 and LAS by +8.5 on average. We also present semi-supervised transfer and learning curve experiments that show ESR provides significant gains over strong cross-lingual-transfer-plus-fine-tuning baselines for modest amounts of label data. These results indicate that ESR is a promising and complementary approach to modeltransfer approaches for cross-lingual parsing. 1 | Improving Low-Resource Cross-lingual Parsing with Expected Statistic Regularization |
d227231142 | This paper reports on the harvesting, analysis, and annotation of 20k documents from 4 different endangered language archives in 280 di erent low-resource languages. The documents are heterogeneous as to their provenance (holding archive, language, geographical area, creator) and internal structure (annotation types, metalanguages), but they have the ELAN-XML format in common. Typical annotations include sentence-level translations, morpheme-segmentation, morpheme-level translations, and parts-of-speech. The ELAN format gives a lot of freedom to document creators, and hence the data set is very heterogeneous. We use regularities in the ELAN format to arrive at a common internal representation of sentences, words, and morphemes, with translations into one or more additional languages. Building upon the paradigm of Linguistic Linked Open Data (LLOD, Chiarcos et al. (2012b)), the document elements receive unique identi ers and are linked to other resources such as Glottolog for languages, Wikidata for semantic concepts, and the Leipzig Glossing Rules list for category abbreviations. We provide an RDF export in the LIGT format (Chiarcos and Ionov(2019)), enabling uniform and interoperable access with some semantic enrichments to a formerly disparate resource type di cult to access. Two use cases (semantic search and colexi cation) are presented to show the viability of the approach. | Modelling and annotating interlinear glossed text from 280 di erent endangered languages as Linked Data with LIGT |
d252847548 | The Dorabella cipher is an encrypted note written by English composer Edward Elgar, which has defied decipherment attempts for more than a century. While most proposed solutions are English texts, we investigate the hypothesis that Dorabella represents enciphered music. We weigh the evidence for and against the hypothesis, devise a simplified music notation, and attempt to reconstruct a melody from the cipher. Our tools are n-gram models of music which we validate on existing music corpora enciphered using monoalphabetic substitution. By applying our methods to Dorabella, we produce a decipherment with musical qualities, which is then transformed via artful composition into a listenable melody. Far from arguing that the end result represents the only true solution, we instead frame the process of decipherment as part of the composition process. | Dorabella Cipher as Musical Inspiration |
d33728600 | This paper is an introdnction to KASSYS, a system that has been designed to extract information from detining statements in natural language. Only hyperonymous detinitions are dealt with here, for which systematic processing has been devised and itnl)lenmnted in tlte initial version of the system. The paper describes how KASSYS buiMs a taxinomie hierarchy by extracting the hyperonyms from these definitions. It also explains the way in which the system can answer closed questions (yes/no), thus enabling the user to check very quickly that a definition has been assimilated correctly. The underlying forrnalism is that of conceptual graphs, with which the reader is assumed to be familiar. | KASSYS: A DEFINITION ACQUISITION SYSTEM IN NATURAL LANGUAGE |
d269066 | We examine the feasibility of harvesting a wide-coverage lexicon of English verbs from the FrameNet semantically annotated corpus, intended for use in a practical natural language understanding (NLU) system. We identify a range of constructions for which current annotation practice leads to problems in deriving appropriate lexical entries, for example imperatives, passives and control, and discuss potential solutions. | Extracting a verb lexicon for deep parsing from FrameNet |
d53591648 | Kanyen'kéha (in English, Mohawk) is an Iroquoian language spoken primarily in Eastern Canada (Ontario, Québec). Classified as endangered, it has only a small number of speakers and very few younger native speakers. Consequently, teachers and courses, teaching materials and software are urgently needed. In the case of software, the polysynthetic nature of Kanyen'kéha means that the number of possible combinations grows exponentially and soon surpasses attempts to capture variant forms by hand. It is in this context that we describe an attempt to produce language teaching materials based on a generative approach. A natural language generation environment (ivi/Vinci) embedded in a web environment (VinciLingua) makes it possible to produce, by rule, variant forms of indefinite complexity. These may be used as models to explore, or as materials to which learners respond. Generated materials may take the form of written text, oral utterances, or images; responses may be typed on a keyboard, gestural (using a mouse) or, to a limited extent, oral. The software also provides complex orthographic, morphological and syntactic analysis of learner productions. We describe the trajectory of development of materials for a suite of four courses on Kanyen'kéha, the first of which will be taught in the fall of 2018. | Natural Language Generation for Polysynthetic Languages: Language Teaching and Learning Software for Kanyen'kéha (Mohawk) |
d70254353 | Grammaires d'unification ~ traits et contr61e des infinitives en fran~ais (Unification Grammars with Features and Control of Infinitives in French) Reviewed by Dominique Estival Universit~ de Gen~ve | |
d233189541 | ||
d236477351 | ||
d23247513 | This paper introduces a toolkit used for the purpose of detecting replacements of different grammatical and semantic structures in ongoing text production logged as a chronological series of computer interaction events (so-called keystroke logs). The specific case we use involves human translations where replacements can be indicative of translator behaviour that leads to specific features of translations that distinguish them from non-translated texts. The toolkit uses a novel CCG chart parser customised so as to recognise grammatical words independently of space and punctuation boundaries. On the basis of the linguistic analysis, structures in different versions of the target text are compared and classified as potential equivalents of the same source text segment by 'equivalence judges'. In that way, replacements of grammatical and semantic structures can be detected. Beyond the specific task at hand the approach will also be useful for the analysis of other types of spaceless text such as Twitter hashtags and texts in agglutinative or spaceless languages like Finnish or Chinese. | Automatic recognition of linguistic replacements in text series generated from keystroke logs |
d218973935 | ||
d14219758 | The Arabic Treebank (ATB), released by the Linguistic Data Consortium, contains multiple annotation files for each source file, due in part to the role of diacritic inclusion in the annotation process. The data is made available in both "vocalized" and "unvocalized" forms, with and without the diacritic marks, respectively. Much parsing work with the ATB has used the unvocalized form, on the basis that it more closely represents the "real-world" situation. We point out some problems with this usage of the unvocalized data and explain why the unvocalized form does not in fact represent "real-world" data. This is due to some aspects of the treebank annotation that to our knowledge have never before been published. | Diacritic Annotation in the Arabic Treebank and Its Impact on Parser Evaluation |
d15574572 | This paper reports MT evaluation experiments that were conducted at the end of year 1 of the EU-funded CoSyne 1 project for three language combinations, considering translations from German, Italian and Dutch into English. We present a comparative evaluation of the MT software developed within the project against four of the leading free webbased MT systems across a range of state-of-the-art automatic evaluation metrics. The data sets from the news domain that were created and used for training purposes and also for this evaluation exercise, which are available to the research community, are also described. The evaluation results for the news domain are very encouraging: the CoSyne MT software consistently beats the rule-based MT systems, and for translations from Italian and Dutch into English in particular the scores given by some of the standard automatic evaluation metrics are not too distant from those obtained by wellestablished statistical online MT systems. 2 | Comparative Evaluation of Research vs. Online MT Systems |
d252847579 | End-to-end task bots are typically learned over a static and usually limited-size corpus. However, when deployed in dynamic, changing, and open environments to interact with users, task bots tend to fail when confronted with data that deviate from the training corpus, i.e., out-ofdistribution samples. In this paper, we study the problem of automatically adapting task bots to changing environments by learning from human-bot interactions with minimum or zero human annotations. We propose SL-AGENT 1 , a novel self-learning framework for building end-to-end task bots. SL-AGENT consists of a dialog model and a pre-trained reward model to predict the quality of an agent response. It enables task bots to automatically adapt to changing environments by learning from the unlabeled human-bot dialog logs accumulated after deployment via reinforcement learning with the incorporated reward model. Experimental results on four well-studied dialog tasks show the effectiveness of SL-AGENT to automatically adapt to changing environments, using both automatic and human evaluations. We will release code and data for further research. | Toward Self-Learning End-to-End Task-Oriented Dialog Systems |
d14430858 | We present SMMR, a scalable sentence scoring method for query-oriented update summarization. Sentences are scored thanks to a criterion combining query relevance and dissimilarity with already read documents (history). As the amount of data in history increases, non-redundancy is prioritized over query-relevance. We show that SMMR achieves promising results on the DUC 2007 update corpus. | Coling 2008: Companion volume -Posters and Demonstrations |
d10362187 | There is significant evidence in the literature that integrating knowledge about multiword expressions can improve shallow parsing accuracy. We present an experimental study to quantify this improvement, focusing on compound nominals, proper names and adjectivenoun constructions. The evaluation set of multiword expressions is derived from Word-Net and the textual data are downloaded from the web. We use a classification method to aid human annotation of output parses. This method allows us to conduct experiments on a large dataset of unannotated data. Experiments show that knowledge about multiword expressions leads to an increase of between 7.5% and 9.5% in accuracy of shallow parsing in sentences containing these multiword expressions. | Can Recognising Multiword Expressions Improve Shallow Parsing? |
d52817936 | We present a novel approach to deciding whether two sentences hold a paraphrase relationship. We employ a generative model that generates a paraphrase of a given sentence, and we use probabilistic inference to reason about whether two sentences share the paraphrase relationship. The model cleanly incorporates both syntax and lexical semantics using quasi-synchronous dependency grammars(Smith and Eisner, 2006). Furthermore, using a product of experts(Hinton, 2002), we combine the model with a complementary logistic regression model based on state-of-the-art lexical overlap features. We evaluate our models on the task of distinguishing true paraphrase pairs from false ones on a standard corpus, giving competitive state-of-the-art performance. | Paraphrase Identification as Probabilistic Quasi-Synchronous Recognition |
d227905424 | ||
d235599170 | ||
d14849127 | Analyzing public opinions towards products, services and social events is an important but challenging task. An accurate sentiment analyzer should take both lexicon-level information and corpus-level information into account. It also needs to exploit the domainspecific knowledge and utilize the common knowledge shared across domains. In addition, we want the algorithm being able to deal with missing labels and learning from incomplete sentiment lexicons. This paper presents a LCCT (Lexicon-based and Corpus-based, Co-Training) model for semi-supervised sentiment classification. The proposed method combines the idea of lexicon-based learning and corpus-based learning in a unified cotraining framework. It is capable of incorporating both domain-specific and domainindependent knowledge. Extensive experiments show that it achieves very competitive classification accuracy, even with a small portion of labeled data. Comparing to state-ofthe-art sentiment classification methods, the LCCT approach exhibits significantly better performances on a variety of datasets in both English and Chinese. | LCCT: A Semi-supervised Model for Sentiment Classification |
d252280427 | As political attitudes have diverged ideologically in the United States, political speech has diverged linguistically. The ever-widening polarization between the US political parties is accelerated by an erosion of mutual understanding between them. We aim to make these communities more comprehensible to each other with a framework that probes community-specific responses to the same survey questions using community language models (COMMUNITYLM). In our framework we identify committed partisan members for each community on Twitter and fine-tune LMs on the tweets authored by them. We then assess the worldviews of the two groups using promptbased probing of their corresponding LMs, with prompts that elicit opinions about public figures and groups surveyed by the American National Election Studies (ANES) 2020 Exploratory Testing Survey. We compare the responses generated by the LMs to the ANES survey results, and find a level of alignment that greatly exceeds several baseline methods. Our work aims to show that we can use community LMs to query the worldview of any group of people given a sufficiently large sample of their social media discussions or media diet. . 2020. oLMpics-on what language model pre-training captures. Transactions of the Association for Computational Linguistics, 8:743-758. | COMMUNITYLM: Probing Partisan Worldviews from Language Models |
d1013580 | One may express favor (or disfavor) towards a target by using positive or negative language. Here for the first time we present a dataset of tweets annotated for whether the tweeter is in favor of or against pre-chosen targets, as well as for sentiment. These targets may or may not be referred to in the tweets, and they may or may not be the target of opinion in the tweets. We develop a simple stance detection system that outperforms all 19 teams that participated in a recent shared task competition on the same dataset (SemEval-2016 Task #6). Additionally, access to both stance and sentiment annotations allows us to conduct several experiments to tease out their interactions. We show that while sentiment features are useful for stance classification, they alone are not sufficient. We also show the impacts of various features on detecting stance and sentiment, respectively. | Detecting Stance in Tweets And Analyzing its Interaction with Sentiment |
d6879481 | In this paper we present the Multilingual All-Words Sense Disambiguation and Entity Linking task. Word Sense Disambiguation (WSD) and Entity Linking (EL) are well-known problems in the Natural Language Processing field and both address the lexical ambiguity of language. Their main difference lies in the kind of meaning inventories that are used: EL uses encyclopedic knowledge, while WSD uses lexicographic information. Our aim with this task is to analyze whether, and if so, how, using a resource that integrates both kinds of inventories (i.e., BabelNet 2.5.1) might enable WSD and EL to be solved by means of similar (even, the same) methods. Moreover, we investigate this task in a multilingual setting and for some specific domains. | SemEval-2015 Task 13: Multilingual All-Words Sense Disambiguation and Entity Linking |
d198147607 | Many Natural Language Processing (NLP) tasks depend on using Named Entities (NEs) that are contained in texts and in external knowledge sources. While this is easy for humans, the present neural methods that rely on learned word embeddings may not perform well for these NLP tasks, especially in the presence of Out-Of-Vocabulary (OOV) or rare NEs. In this paper, we propose a solution for this problem, and present empirical evaluations on: a) a structured Question-Answering task, b) three related Goal-Oriented dialog tasks, and c) a Reading-Comprehension task 1 , which show that the proposed method can be effective in dealing with both in-vocabulary and OOV NEs. * Equal Contribution † This work was done when the author was at IBM Research, NY. 1 We create extended versions of dialog bAbI tasks 1,2 and 4 and OOV versions of the CBT test sethttps: //github.com/IBM/ne-table-datasets/ | Lazaros Polymenakos † |
d6215504 | What is the role of textual features above the sentence level in advancing the machine translation of literature? This paper examines how referential cohesion is expressed in literary and non-literary texts and how this cohesion affects translation. We first show in a corpus study on English that literary texts use more dense reference chains to express greater referential cohesion than news. We then compare the referential cohesion of machine versus human translations of Chinese literature and news. While human translators capture the greater referential cohesion of literature, Google translations perform less well at capturing literary cohesion. Our results suggest that incorporating discourse features above the sentence level is an important direction for MT research if it is to be applied to literature. | Towards a Literary Machine Translation: The Role of Referential Cohesion |
d219303250 | ||
d250390923 | We present the findings of SemEval-2022 Task 11 on Multilingual Complex Named Entity Recognition MULTICONER. 1 Divided into 13 tracks, the task focused on methods to identify complex named entities (like media titles, products, and groups) in 11 languages in both monolingual and multi-lingual scenarios. Eleven tracks were for building monolingual NER models for individual languages, one track focused on multilingual models able to work on all languages, and the last track featured code-mixed texts within any of these languages. The task used the MULTICONER dataset, composed of 2.3 million instances in Bangla, Chinese, Dutch, English, Farsi, German, Hindi, Korean, Russian, Spanish, and Turkish. Results showed that methods fusing external knowledge into transformer models achieved the best performance. The largest gains were on the Creative Work and Group entity classes, which are still challenging even with external knowledge. MULTICONER was one of the most popular tasks in SemEval-2022 and it attracted 377 participants during the practice phase. The final test phase had 236 participants, and 55 teams submitted their systems. * These authors contributed equally to this work. 1 https://multiconer.github.io/ | SemEval-2022 Task 11: Multilingual Complex Named Entity Recognition (MultiCoNER) |
d227230584 | ||
d52010094 | The rapid growth of documents across the web has necessitated finding means of discarding redundant documents and retaining novel ones. Capturing redundancy is challenging as it may involve investigating at a deep semantic level. Techniques for detecting such semantic redundancy at the document level are scarce. In this work we propose a deep Convolutional Neural Network (CNN) based model to classify a document as novel or redundant with respect to a set of relevant documents already seen by the system. The system is simple and does not require manual feature engineering. Our novel scheme encodes relevant and relative information from both source and target texts to generate an intermediate representation for which we coin the name Relative Document Vector (RDV). The proposed method outperforms the existing benchmark on two document-level novelty detection datasets by a margin of ∼5% in terms of accuracy. We further demonstrate the effectiveness of our approach on a standard paraphrase detection dataset where the paraphrased passages closely resembles semantically redundant documents. This work is licenced under a Creative Commons Attribution 4.0 International Licence.Licence details: | Novelty Goes Deep. A Deep Neural Solution To Document Level Novelty Detection |
d2988658 | In this paper, we introduce an automatic method for classifying a given question using broad semantic categories in an existing lexical database (i.e., WordNet) as the class tagset. For this, we also constructed a large scale entity supersense database that contains over 1.5 million entities to the 25 WordNet lexicographer's files (supersenses) from titles of Wikipedia entry. To show the usefulness of our work, we implement a simple redundancy-based system that takes the advantage of the large scale semantic database to perform question classification and named entity classification for open domain question answering. Experimental results show that the proposed method outperform the baseline of not using question classification. | Minimally Supervised Question Classification and Answering based on WordNet and Wikipedia |
d6540554 | Recent work has shown success in using continuous word embeddings learned from unlabeled data as features to improve supervised NLP systems, which is regarded as a simple semi-supervised learning mechanism. However, fundamental problems on effectively incorporating the word embedding features within the framework of linear models remain. In this study, we investigate and analyze three different approaches, including a new proposed distributional prototype approach, for utilizing the embedding features. The presented approaches can be integrated into most of the classical linear models in NLP. Experiments on the task of named entity recognition show that each of the proposed approaches can better utilize the word embedding features, among which the distributional prototype approach performs the best. Moreover, the combination of the approaches provides additive improvements, outperforming the dense and continuous embedding features by nearly 2 points of F1 score. | Revisiting Embedding Features for Simple Semi-supervised Learning |
d229365791 | This paper describes the University of Maryland's submissions to the WMT20 Shared Task on Chat Translation. We focus on translating agent-side utterances from English to German. We started from an off-the-shelf BPEbased standard transformer model trained with WMT17 news and fine-tuned it with the provided in-domain training data. In addition, we augment the training set with its best matches in the WMT19 news dataset. Our primary submission uses a standard Transformer, while our contrastive submissions use multi-encoder Transformers to attend to previous utterances. Our primary submission achieves 56.7 BLEU on the agent side (en→de), outperforming a baseline system provided by the task organizers by more than 13 BLEU points. Moreover, according to an evaluation on a set of carefullydesigned examples, the multi-encoder architecture is able to generate more coherent translations. | The University of Maryland's Submissions to the WMT20 Chat Translation Task: Searching for More Data to Adapt Discourse-Aware Neural Machine Translation |
d13974252 | Morphological analysis and disambiguation are crucial stages in a variety of natural language processing applications, especially when languages with complex morphology are concerned. We present a system which disambiguates the output of a morphological analyzer for Hebrew. It consists of several simple classifiers and a module which combines them under linguistically motivated constraints. We investigate a number of techniques for combining the predictions of the classifiers. Our best result, 91.44% accuracy, reflects a 25% reduction in error rate compared with the previous state of the art. | Morphological Disambiguation of Hebrew: A Case Study in Classifier Combination |
d12879055 | An important part of the development of any machine translation system is the creation of lexical resources. We describe an analysis of the dictionary development workflow and supporting tools currently in use and under development at Logos. This workflow identifies the component processes of: setting goals, locating and acquiring lexical resources, transforming the resources to a common format, classifying and routing entries for special processing, importing entries, and verifying their adequacy in translation. Our approach has been to emphasize the tools necessary to support increased automation and use of resources available in electronic formats, in the context of a systematic workflow design. | Dictionary Development Workflow for MT: Design and Management |
d16021345 | AutoCor is a method for the automatic acquisition and classification of corpora of documents in closely-related languages. It is an extension and enhancement of CorpusBuilder, a system that automatically builds specific minority language corpora from a closed corpus, since some Tagalog documents retrieved by CorpusBuilder are actually documents in other closely-related Philippine languages. AutoCor used the query generation method odds ratio, and introduced the concept of common word pruning to differentiate between documents of closely-related Philippine languages and Tagalog. The performance of the system using with and without pruning are compared, and common word pruning was found to improve the precision of the system. | AutoCor: A Query Based Automatic Acquisition of Corpora of Closely-related Languages * |
d46530512 | Word embeddings were used for the extraction of hyponymy relation in several approaches, but also it was recently shown that they should not work, in fact. In our work we verified both claims using a very large wordnet of Polish as a gold standard for lexico-semantic relations and word embeddings extracted from a very large corpus of Polish. We showed that a hyponymy extraction method based on linear regression classifiers trained on clusters of vectors can be successfully applied on large scale. We presented also a possible explanation for contradictory findings in the literature. Moreover, in order to show the feasibility of the method we extended it to the recognition of meronymy. | Recognition of Hyponymy and Meronymy Relations in Word Embeddings for Polish |
d17189799 | Support verb constructions (SVC), are verb-noun complexes which play a role in many natural language processing (NLP) tasks, such as Machine Translation (MT). They can be paraphrased with a full verb, preserving its meaning, improving at the same time the MT raw output. In this paper, we discuss the creation of linguistic resources namely a set of dictionaries and rules that can identify and paraphrase Italian SVCs. We propose a paraphrasing computational method that is based on open-source tools and data such as NooJ linguistic environment and OpenLogos MT system. We focus on pre-processing the data that will be machine translated, but our methodology can also be applied in other fields in NLP. Our results show that linguistic knowledge constitutes a 95.5% precision rate in identifying SVC and an 88.8% precision rate in paraphrasing SVCs into full verbs. | Paraphrasing of Italian Support Verb Constructions based on Lexical and Grammatical Resources |
d235097525 | This paper describes our submission for the LongSumm task in SDP 2021. We propose a method for incorporating sentence embeddings produced by deep language models into extractive summarization techniques based on graph centrality in an unsupervised manner. The proposed method is simple, fast, can summarize any document of any size and can satisfy any length constraints for the summaries produced. The method offers competitive performance to more sophisticated supervised methods and can serve as a proxy for abstractive summarization techniques. * Please send correspondence to juan. | Unsupervised document summarization using pre-trained sentence embeddings and graph centrality |
d61428408 | This paper describes VERTa's submission to the 2015 EMNLP Workshop on Statistical Machine Translation. VERTa is a linguistically-motivated metric that combines linguistic features at different levels. In this paper, VERTa is described briefly, as well as the three versions submitted to the workshop: VERTa-70Adeq30Flu, VERTa-EQ and VERTa-W. Finally, the experiments conducted with the WMT14 data are reported and some conclusions are drawn. | VERTa: a Linguistically-motivated Metric at the WMT15 Metrics Task |
d14594 | This paper describes our systems submitted to the target-dependent sentiment polarity classification subtask in aspect based sentiment analysis (ABSA) task (i.e., Task 12) in Se-mEval 2015. To settle this problem, we extracted several effective features from three sequential sentences, including sentiment lexicon, linguistic and domain specific features. Then we employed these features to construct classifiers using supervised classification algorithm. In laptop domain, our systems ranked 2nd out of 6 constrained submissions and 2nd out of 7 unconstrained submissions. In restaurant domain, the rankings are 5th out of 6 and 2nd out of 8 respectively. | ECNU: Extracting Effective Features from Multiple Sequential Sentences for Target-dependent Sentiment Analysis in Reviews |
d258463919 | While prospect theory(Kahneman and Tversky, 1979) has been widely used as a descriptive theory to explain various phenomena such as insurance and the relation between spending and saving, no studies have been found to look at prospect theory from linguistic prospective to investigate near synonymous verbs. This study investigates nearsynonymous verbs in Mandarin Chinese under the framework of prospect theory.Using the large-scale Gigaword2 Corpus, we examined how verbal transaction reflects gain and loss frame in terms of human decision. It is observed that mai3 'buy', mai4 'sell' and zu1 'rent' have default gain-loss frame encoded in the lexical meaning. Because of the topic-prominent typological feature of Mandarin, gain and loss frame have close relation with information structure. By using highlighting and omission of argument, the gainloss frame can be emphasized. Our current study focuses on near synonymous verbs in Mandarin Chinese. It is hoped that more near synonymous verbs in Mandarin Chinese and in other languages can be investigated under the framework of prospect theory. | Gain-framed Buying or Loss-framed Selling? The Analysis of Near Synonyms in Mandarin in Prospect Theory |
d14365086 | We describe FSS-TimEx, a module for the recognition and normalization of temporal expressions we submitted to Task A and B of the TempEval-3 challenge. FSS-TimEx was developed as part of a multilingual event extraction system, Nexus, which runs on top of the EMM news processing engine. It consists of finite-state rule cascades, using minimalistic text processing stages and simple heuristics to model the relations between events and temporal expressions. Although FSS-TimEx is already deployed within an IE application in the medical domain, we found it useful to customize its output to the TimeML standard in order to have an independent performance measure and guide further developments. | FSS-TimEx for TempEval-3: Extracting Temporal Information from Text |
d17561814 | Even syntactically correct sentences are perceived as awkward if they do not contain correct punctuation. Still, the problem of automatic generation of punctuation marks has been largely neglected for a long time. We present a novel model that introduces punctuation marks into raw text material with transition-based algorithm using LSTMs. Unlike the state-of-the-art approaches, our model is language-independent and also neutral with respect to the intended use of the punctuation. Multilingual experiments show that it achieves high accuracy on the full range of punctuation marks across languages. | A Neural Network Architecture for Multilingual Punctuation Generation |
d248811462 | Named entity recognition (NER) is a fundamental and important task in NLP, aiming at identifying named entities (NEs) from free text. Recently, since the multi-head attention mechanism applied in the Transformer model can effectively capture longer contextual information, Transformer-based models have become the mainstream methods and have achieved significant performance in this task. Unfortunately, although these models can capture effective global context information, they are still limited in the local feature and position information extraction, which is critical in NER. In this paper, to address this limitation, we propose a novel Hero-Gang Neural structure (HGN), including the Hero and Gang module, to leverage both global and local information to promote NER. Specifically, the Hero module is composed of a Transformer-based encoder to maintain the advantage of the self-attention mechanism, and the Gang module utilizes a multi-window recurrent module to extract local features and position information under the guidance of the Hero module. Afterward, the proposed multiwindow attention effectively combines global information and multiple local features for predicting entity labels. Experimental results on several benchmark datasets demonstrate the effectiveness of our proposed model. 1 | Hero-Gang Neural Model For Named Entity Recognition |
d1673975 | We present MIA, a data marketplace which enables massive parallel processing of data from the Web. End users can combine both text mining and database operators in a structured query language called MIAQL. MIA offers many cost savings through sharing text data, annotations, built-in analytical functions and third party text mining applications. Our demonstration showcases MIAQL and its execution on the platform for the example of analyzing political campaigns. | A Marketplace for Web Scale Analytics and Text Annotation Services |
d41952796 | Word vectors with varying dimensionalities and produced by different algorithms have been extensively used in NLP. The corpora that the algorithms are trained on can contain either natural language text (e.g. Wikipedia or newswire articles) or artificially-generated pseudo corpora due to natural data sparseness.We exploit Lexical Chain based templates over Knowledge Graph for generating pseudo-corpora with controlled linguistic value. These corpora are then used for learning word embeddings. A number of experiments have been conducted over the following test sets: WordSim353 Similarity, WordSim353 Relatedness and SimLex-999.The results show that, on the one hand, the incorporation of many-relation lexical chains improves results, but on the other hand, unrestricted-length chains remain difficult to handle with respect to their huge quantity. | Towards Lexical Chains for Knowledge-Graph-based Word Embeddings |
d250279955 | We describe our two-stage system for the Multilingual Information Access (MIA) 2022 Shared Task on Cross-Lingual Open-Retrieval Question Answering. The first stage consists of multilingual passage retrieval with a hybrid dense and sparse retrieval strategy. The second stage consists of a reader which outputs the answer from the top passages returned by the first stage. We show the efficacy of using entity representations, sparse retrieval signals to help dense retrieval, and Fusion-in-Decoder. On the development set, we obtain 43.46 F1 on XOR-TyDi QA and 21.99 F1 on MKQA, for an average F1 score of 32.73. On the test set, we obtain 40.93 F1 on XOR-TyDi QA and 22.29 F1 on MKQA, for an average F1 score of 31.61. We improve over the official baseline by over 4 F1 points on both the development and test sets. 1 | MIA 2022 Shared Task Submission: Leveraging Entity Representations, Dense-Sparse Hybrids, and Fusion-in-Decoder for Cross-Lingual Question Answering |
d203690908 | ||
d18927627 | Search personalization that considers the social dimension of the web has attracted a significant volume of research in recent years. A user profile is usually needed to represent a user's interests in order to tailor future searches. Previous research has typically constructed a profile solely from a user's usage information. When the user has only limited activities in the system, the effect of the user profile on search is also constrained. This research addresses the setting where a user has only a limited amount of usage information. We build enhanced user profiles from a set of annotations and resources that users have marked, together with an external knowledge base constructed according to usage histories. We present two probabilistic latent topic models to simultaneously incorporate social annotations, documents and the external knowledge base. Our web search strategy is achieved using personalized social query expansion. We introduce a topical query expansion model to enhance the search by utilizing individual user profiles. The proposed approaches have been intensively evaluated on a large public social annotation dataset. Results show that our models significantly outperformed existing personalized query expansion methods which use user profiles solely built from past usage information in personalized search. | Enhanced Personalized Search using Social Data |
d259137871 | Currently, the reduction in the parameter scale of large-scale pre-trained language models (PLMs) through knowledge distillation has greatly facilitated their widespread deployment on various devices. However, the deployment of knowledge distillation systems faces great challenges in real-world industrial-strength applications, which require the use of complex distillation methods on even larger-scale PLMs (over 10B), limited by memory on GPUs and the switching of methods. To overcome these challenges, we propose GKD, a general knowledge distillation framework that supports distillation on larger-scale PLMs using various distillation methods. With GKD, developers can build larger distillation models on memorylimited GPUs and easily switch and combine different distillation methods within a single framework. Experimental results show that GKD can support the distillation of at least 100B-scale PLMs and 25 mainstream methods on 8 NVIDIA A100 (40GB) GPUs. 1 | GKD: A General Knowledge Distillation Framework for Large-scale Pre-trained Language Model |
d15575006 | Multiple approaches to grab comparable data from the Web have been developed up to date. Nevertheless, coming out with a high-quality comparable corpus of a specific topic is not straightforward. We present a model for the automatic extraction of comparable texts in multiple languages and on specific topics from Wikipedia. In order to prove the value of the model, we automatically extract parallel sentences from the comparable collections and use them to train statistical machine translation engines for specific domains. Our experiments on the English-Spanish pair in the domains of Computer Science, Science, and Sports show that our in-domain translator performs significantly better than a generic one when translating in-domain Wikipedia articles. Moreover, we show that these corpora can help when translating out-of-domain texts. | A Factory of Comparable Corpora from Wikipedia |
d219307917 | ||
d820475 | Speaker's intention prediction modules can be widely used as a pre-processor for reducing the search space of an automatic speech recognizer. They also can be used as a preprocessor for generating a proper sentence in a dialogue system. We propose a statistical model to predict speakers' intentions by using multi-level features. Using the multi-level features (morpheme-level features, discourselevel features, and domain knowledge-level features), the proposed model predicts speakers' intentions that may be implicated in next utterances. In the experiments, the proposed model showed better performances (about 29% higher accuracies) than the previous model. Based on the experiments, we found that the proposed multi-level features are very effective in speaker's intention prediction. | Speakers' Intention Prediction Using Statistics of Multi-level Features in a Schedule Management Domain |
d252118507 | Discourse-aware techniques, including entityaware approaches, play a crucial role in summarization. In this paper, we propose an entity-based SpanCopy mechanism to tackle the entity-level factual inconsistency problem in abstractive summarization, i.e. reducing the mismatched entities between the generated summaries and the source documents. Complemented by a Global Relevance component to identify summary-worthy entities, our approach demonstrates improved factual consistency while preserving saliency on four summarization datasets, contributing to the effective application of discourse-aware methods summarization tasks. 1 1 The code is available at https://github.com/ Wendy-Xiao/Entity-based-SpanCopy | Entity-based SpanCopy for Abstractive Summarization to Improve the Factual Consistency |
d12591723 | This article evaluates the integration of data extracted from a syntactic lexicon, namely the Lexicon-Grammar, into several probabilistic parsers for French. We show that by modifying the Part-of-Speech tags of verbs and verbal nouns of a treebank, we obtain accurate performances with a parser based on Probabilistic Context-Free Grammars(Petrov et al., 2006)and a discriminative parser based on a reranking algorithm(Charniak and Johnson, 2005). | Integration of Data from a Syntactic Lexicon into Generative and Discriminative Probabilistic Parsers |
d12353337 | This paper presents two systems for textual entailment, both employing decision trees as a supervised learning algorithm. The first one is based primarily on the concept of lexical overlap, considering a bag of words similarity overlap measure to form a mapping of terms in the hypothesis to the source text. The second system is a lexicosemantic matching between the text and the hypothesis that attempts an alignment between chunks in the hypothesis and chunks in the text, and a representation of the text and hypothesis as two dependency graphs. Their performances are compared and their positive and negative aspects are analyzed. | Textual Entailment Through Extended Lexical Overlap and Lexico-Semantic Matching |
d1368790 | Metaphors pervade our language because they are elastic enough to allow a speaker to express an affective viewpoint on a topic without committing to a specific meaning. This balance of expressiveness and indeterminism means that metaphors are just as useful for eliciting information as they are for conveying information. We explore here, via a demonstration of a system for metaphor interpretation and generation called Metaphor Magnet, the practical uses of metaphor as a basis for formulating affective information queries. We also consider the kinds of deep and shallow stereotypical knowledge that are needed for such a system, and demonstrate how they can be acquired from corpora and the web. | Specifying Viewpoint and Information Need with Affective Metaphors A System Demonstration of the Metaphor Magnet Web App/Service |
d257920051 | In an information-seeking conversation, a user may ask questions that are under-specified or unanswerable. An ideal agent would interact by initiating different response types according to the available knowledge sources. However, most current studies either fail to or artificially incorporate such agent-side initiative. This work presents INSCIT, a dataset for Information-Seeking Conversations with mixed-initiative Interactions. It contains 4.7K user-agent turns from 805 human-human conversations where the agent searches over Wikipedia and either directly answers, asks for clarification, or provides relevant information to address user queries. The data supports two subtasks, evidence passage identification and response generation, as well as a human evaluation protocol to assess model performance. We report results of two systems based on state-of-the-art models of conversational knowledge identification and opendomain question answering. Both systems significantly underperform humans, suggesting ample room for improvement in future studies. 1 | INSCIT: Information-Seeking Conversations with Mixed-Initiative Interactions |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.