text stringlengths 4 222k | label int64 0 4 |
|---|---|
Contextual word representations (CWRs) from pretrained language models have improved many NLP systems. Such language models include BERT and ELMo , which are conventionally "pretrained" on large unlabeled datasets before their internal representations are "finetuned" during supervised training on downstream tasks like ... | 0 |
Named Entity Recognition (NER) is a task that seeks to locate and classify entities ('atomic elements') in a text into predefined categories such as the names of persons, organizations, locations, expressions of times, quantities, etc. It can be viewed as a two stage process:For example, if "Mahatma Gandhi" is a named ... | 0 |
One of the most important steps for building accurate QA systems is the selection/reranking of answer passage (AP) candidates typically provided by a search engine. This task requires the automatic learning of a ranking function, which pushes the correct answer passages (i.e., containing the answer to the question) hig... | 0 |
Although task-oriented dialogue systems have been actively researched (Hirshman, 1989; Ferguson et al., 1996; Nakano et al., 1999; Walker et al., 2002) , recently non-task-oriented functions are starting to attract attention, and systems without a specific task that deal with more casual dialogues, such as chats, are b... | 0 |
Arabic is a morphologically rich and complex language, which presents significant challenges for natural language processing and its applications. It is the official language in 22 countries spoken by more than 350 million people around the world 1 . Moreover, the Arabic language exists in a state of diglossia where th... | 0 |
Wikipedia, despite initial scepticism, is an incredibly robust semantic resource. Armed with a shared set of standards, legions of volunteers make positive changes to the pages of this vast encyclopaedia every day. Some of these editors may be casual -perhaps noticing an error in a page they were reading and being moti... | 0 |
A concept or a sense, s, can be defined as the meaning of a word or a multiword expression. A concept s can be linguistically realised by more than one word while at the same time a word w can be the linguistic realisation of more than one concept. Given a set of concepts S, taxonomy learning is the task of hierarchica... | 0 |
Walenty (Walenty, 2016) is a comprehensive valence dictionary of Polish developed at the Institute of Computer Science, Polish Academy of Sciences (ICS PAS) (Przepiórkowski et al., 2014a; Przepiórkowski et al., 2014b) . The dictionary is meant to be both human-and machinereadable; in particular, it is being employed by... | 0 |
Linguistically syntax-based statistical machine translation models have made promising progress in recent years. By incorporating the syntactic annotations of parse trees from both or either side(s) of the bitext, they are believed better than phrasebased counterparts in reorderings. Depending on the type of input, the... | 0 |
Linguistic resources are of fundamental importance for conducting linguistic research. Some of these resources are originally created in handwritten form but need to be digitised in order to be efficiently exploited. In the absence of automatic handwriting recognition systems capable of handling in a satisfactory fashi... | 0 |
Text-oriented methods and software tools have come to be of primary interest to the NLP community. However, existing tools for natural language processing (NLP) and machine translation (MT) corpus-based research are typically embedded in large, non-adaptable systems which are fundamentally incompatible. Little effort h... | 0 |
Natural language processing research relies heavily on the availability of textual corpora annotated with various levels of syntactic and semantic information such as treebanks (Marcus et al., 1993) or propbanks (Palmer et al., 2005) . However, the manual creation of such resources is known to be highly costly and ther... | 0 |
Product reviews written by the users on e-commerce websites are useful to get the feedback about the products and provide valuable insights to the user for making the buying decision. The product reviews available on different e-commerce websites are mainly in English language. India is a multilingual country with grea... | 0 |
Document embedding is important for many NLP tasks such as document classification (e.g., essay scoring and sentiment classification) (Le and Mikolov, 2014; Liu et al., 2017; Wu et al., 2018; Tang et al., 2015) and summarization. While embedding approaches can be supervised, semi-supervised and unsupervised, recent stu... | 0 |
China and Japan are producing a large amount of patents in their respective languages. Making Chinese patents available in Japanese, and Japanese patents in Chinese is an important task for increasing economical development in Asia and international world. The translation of patents is a key issue that should be helped... | 0 |
The goal of Relation Extraction (RE) is to find and classify instances of certain relations in raw text. We denote a binary relation instance, i.e. a relation instance with two arguments, with a tuple x = (s, e 1 , e 2 , r), where s = [w 0 • • • w n ] is a sequence of sentence tokens, e 1 , e 2 are entity mentions with... | 0 |
Quality estimation (QE) evaluates the quality of machine translation output without human reference translations (Blatz et al., 2004) . It has a wide range of applications in post-editing and quality control for machine translation.We participated in all tasks and language pairs at the WMT 2020 QE shared task 2 , inclu... | 0 |
Natural-Language Generation (NLG) systems are usually evaluated quantitatively, by measuring impact on task performance, human opinions on Likert-like scales, and/or similarity to a goldstandard corpus. While such evaluations are essential, we believe there is also a role for qualitative evaluations, especially when th... | 0 |
Fine-grained entity typing (FET) is the task of classifying named entity mentions in a sentence over the given class set (typically a hierarchical class structure as shown in Fig. 1 . FET serves as an important component in many down-stream NLP applications, e.g., relation extraction (Liu et al., 2014) , entity linking... | 0 |
Attention towards sentiment analysis has been flourishing over the last two decades because of the immense popularity of social media. The phenomenal rise in blogging trend is observed in health communities such as medical forums which are swamped by millions of users (many of whom are patients) seeking for health-rela... | 0 |
The goat of using "high level" knowledge sources in recognizing continuous speech is to reduce the hypotheses space generated by acoustic-phonetic analysis (and possibly to implement an interpretation of the utterance) (see for instance Walker 1976 , Stringa 1988 . Deeodifying the vocal signal is a process that must ta... | 0 |
A natural language interface to a computer database provides users with the capability of obtaining information stored in the database by querying the system in a natural language (NL) . With a natural language as a means of communication with a computer system, the users can make a question or a statement in the way t... | 0 |
The Perceptron training algorithm (Rosenblatt, 1958; Freund and Schapire, 1999; Collins, 2002) is widely applied in the Natural Language Processing community for learning complex structured models. The non-probabilistic nature of the perceptron parameters makes it possible to incorporate arbitrary features without the ... | 0 |
Word embeddings (Bengio et al., 2003; Mikolov et al., 2013; Pennington et al., 2014) are the current state of art techniques to form semantic representations of words based on their contexts. They have been successfully used in various downstream tasks such as text classification, text generation, etc. Building on word... | 0 |
Current technology makes the automated capture, storage, indexing, and categorization of broadcast news feasible allowing for the development of computational systems that provide for the intelligent browsing and retrieval of news stories [Maybury, Merlino & Morey '97; Kubula, et al., '00]. To be effective, such system... | 0 |
emphasised that natural language input unfolds in time and therefore, recurrent networks which can accept a sequence of input patterns are the preferred choice for many connectionist natural language processing tasks. In recurrent networks, knowledge is represented in activation patterns over hidden units and revealed ... | 0 |
Once statistical translation models are trained, a decoding approach determines what translations are finally selected. Two parallel lines of research have shown consistent improvements over the standard max-derivation decoding objective, which selects the highest probability derivation. Consensus decoding procedures s... | 0 |
In this paper we describe our participation with a feature-based supervised system to the SemEval-2013 task #7: The Joint Student Response Analysis and 8th Recognizing Textual Entailment Challenge (Dzikovska et al., 2013) . The goal of our participation is to build a generic system that is robust enough across domains ... | 0 |
In recent years, the NLG community has seen a substantial number of studies to evaluate Generation of Referring Expressions (GRE) algorithms, but it is still far from clear what would constitute an optimal evaluation method. Two limitations stand out in the bulk of existing work. Firstly, most existing evaluations are ... | 0 |
Les systèmes d'apprentissage d'une langue seconde assisté par ordinateur tentent d'évaluer automatiquement la prononciation pour aider les apprenants. Pour l'évaluation au niveau segmental, une approche standard consiste à attribuer un score de prononciation à chaque réalisation de phone (Eskenazi, 2009) . Deux grands ... | 0 |
Transformer-based models (Vaswani et al., 2017) like BERT (Devlin et al., 2019) , XLNet (Yang et al., 2019) , RoBERTa , and T5 (Raffel et al., 2019) have pushed the state of the art on an impressive array of NLP tasks. Overparameterized transformers are known to be unversal approximators (Yun et al., 2020 ), suggesting... | 0 |
Millions of internet users send, post, and share text messages every day via various platforms, such as online social networking service (Twitter, Facebook, etc.) , instant messaging service (WhatsApp, Telegram, etc.), online blogs, and forums. One of them, Twitter, has become popular among computational linguistics an... | 0 |
In statistical machine translation, the smallest unit is usually the word, defined as a token delimited by spaces. Given a parallel corpus of source and target text, the training procedure first builds a word alignment, then extracts phrase pairs from this word alignment. However, in some languages (e.g., Chinese) ther... | 0 |
Task-oriented dialog (TOD) systems enable a user to use natural language directions to complete specific tasks. Recently, such systems have been successfully deployed in smart applications such as Amazon's Echo and Spotify's Car Thing.There are several components that are critical to the performance of a TOD system. Th... | 0 |
For some years, the relevance of formal results in grammatical inference to the empirical question of first language acquisition by infant children has been recognised (Wexler and Culicover, 1980) . Unfortunately, for many researchers, with a few notable exceptions (Abe, 1988) , this begins and ends with Gold's famous ... | 0 |
Generalization, the ability to extend patterns from known to unknown items, is a critical part of morphological competence. Morphological systems, both human and machine, must be able to recognize and produce novel items as new words are encountered. Every learner, every speaker, and any system intended for general use... | 0 |
As the Chinese text comes without word delimiters, the Chinese word segmentation becomes a necessary step towards further syntactic analysis. With the evolving of statistical word segmentation techniques (Peng et al., 2004; Kiat Low et al., 2005; Zhang and Clark, 2008) , some of the state-of-the-art systems (Sun, 2011;... | 0 |
What role does linguistics play in statistical language learning? "None at all!" might be the answer, if we ask hard-core speech-recognition professionals. But even the most nonlinguistic language model, for example a statistic word bigram model, actually relies on key concepts integral to virtually all linguistic theo... | 0 |
Ideally, college students would complete assigned readings before class, allowing professors to lean on that shared knowledge, extending and deepening understanding rather than reteaching the textbook context during the class. However, there have been a number of studies showing that when student work is not directly c... | 0 |
NLP systems have been known to learn spurious patterns from data to achieve high accuracy on test sets (Goyal et al., 2017; Gururangan et al., 2018; Glockner et al., 2018; Tsuchiya, 2018; Geva et al., 2019) . Evaluating models on static benchmarks and on test sets that have a similar distribution to the training data h... | 0 |
The field of machine translation has changed remarkably little since its earliest days in the fifties. So far, useful machine translation could only obtained in very restricted domain. We believe one of the problems of traditional machine translation lies in how it deals with unit of translation. Normally Rule-Based Ma... | 0 |
One of the most exciting and promising areas of speech research is large-vocabulary continuous speech recognition. A myriad of applications await a good speech recognizer. However, while many reasonable recognizers exist today, they are impractical and inflexible due to the tedious process of configuring a recognizer. ... | 0 |
Relation extraction aims to identify and categorize the semantic relation between two entity mentions based on the contexts within the sentence. Supervised learning approaches have shown to be effective on this task. However, as relation extraction highly depends on information about entities and their contexts, a supe... | 0 |
We present an integrated environment for morphological analysis of (written) word-forms of the Slovene language. The language belongs to the Slavic family of languages, but exhibits some very idiosyncratic properties (e.g. having also a "dual" number and very rich inflection). Our project of writing a morphological ana... | 0 |
Collaborative grounding is shaped by constraints that are not explicit in successful dialog turns. These constraints combine information from world causal relations, the task under discussion, the communicative intents of the dialog partners, and much else besides. They are used to negotiate meaning, to decide which be... | 0 |
Detecting toxic posts on social network sites is a crucial task for social media moderators in order to keep a clean and friendly space for online discussion. To identify whether a comment or post is toxic or not, social network administrators often read the whole comment or post. However, with a large number of length... | 0 |
Our understanding of the Oceanic languages of Melanesia has so far been based mostly on descriptive accounts rather than primary data, since no documentation existed until recently. For some of these languages, high-quality corpora have now become available, but their exploration is still in its infancy.In our MelaTAMP... | 0 |
We set this work in the context of argumentation, where a central aim is to reason from premises to a conclusion using a rule, e.g. the reasoning pattern Modus Ponens is an argument in classic Propositional Logic. Recent efforts to formalise and instantiate "abstract" argumentation have focused on reasoning with defeas... | 0 |
As a long-standing central problem in natural language processing (NLP), dependency parsing has been dominated by data-driven approaches for decades. The foundation of data-driven parsing is the availability and scale of annotated training data (i.e., treebanks). Numerous efforts have been made towards the construction... | 0 |
Automatic text summarization has been a field of active research in recent years. While most methods are extractive, the implementation details differ considerably depending on the goals of a summarization system. Indeed, the intended use of the summaries may help significantly to adapt a particular summarization appro... | 0 |
In argument similes are like songs in love; they describe much, but prove nothing.Comparisons are fundamental linguistic devices that express the likeness of two things-be it entities, concepts or ideas. Given that their working principle is to emphasize the relation between the shared properties of two arguments (Bred... | 0 |
Over the past several years, a significant consensus has emerged over the definition of the realisation task, through the development of realisers such as REALPRO (Lavoie and Rambow, 1997) , ALETH-GEN (Coch, 1996) , KPML (Bateman, 1997) , FUF/SURGE (Elhadad and Robin, 1996) , HALO-GEN (Langkilde, 2000) , YAG (McRoy et ... | 0 |
Automatic post-editors (APEs) seek to perform the same task as human post-editors: correcting errors in text produced by machine translation (MT) systems. APEs have been used to target a variety of different types of MT errors, from determiner selection (Knight and Chander, 1994) to grammatical agreement (Mareček et al... | 0 |
Social media such as Twitter, Facebook, Tumblr and online forums provide a platform for people seeking social support around various psychological and health problems. Analysis of social media posts can reveal different characteristics about the user, including their health and well-being (Paul and Dredze, 2011) . Info... | 0 |
Multiword Expressions (MWEs) constitute a challenging phraseological phenomenon for Natural Language Processing (NLP). They are formed by more than one word, but the whole expression has to be taken into account in order to understand its meaning (Sag et al., 2002) . They are very frequent in natural language, but thei... | 0 |
The concept of a sentence in written or spoken texts is important in several Natural Language Processing (NLP) tasks, such as morpho-syntactic analysis Finger 2010, Fonseca and , sentiment analysis [Brum et al. 2016] , and speech processing [Mendonça et al. 2014] , among others. However, punctuation marks that constitu... | 0 |
Sense annotated corpora are important resources in NLP as they can be used as training data (e.g., for word sense disambiguation (WSD) or semantic role labelling) or as sources for the acquisition of lexical information (e.g., selectional preference information). Typically, a particular sense inventory from a lexical r... | 0 |
One of the most widely used techniques for machine-aided human translation of the last decade is without doubts a method of human translation supported by a translation memory. This technique can substantially speed up the translation process especially when dealing with large amount of repeated translations, e.g. in t... | 0 |
In (Bouayad-Agha et al., 2012) , we presented the NLG challenge of content selection from semantic web data. The task to perform was described as follows: given a set of RDF triples containing facts about a celebrity, select those triples that are reflected in the target text (i.e., a short biography about that celebri... | 0 |
Current methods for large-scale information extraction take advantage of unstructured text available from either Web documents (Banko et al., 2007; Snow et al., 2006) or, more recently, logs of Web search queries (Paşca, 2007) to acquire useful knowledge with minimal supervision. Given a manually-specified target attri... | 0 |
In psychiatry the Mini-International Neuropsychiatric Interview (M.I.N.I.) is a short structured diagnostic interview for psychiatric disorders (Sheehan et al., 1998) . In Germany it is, among others, used for diagnosing Arabic speaking refugees. Here, the language barrier is an obvious one, that is normally overcome w... | 0 |
Data-to-text NLG is the computational task which aims to generate text from non-linguistic data (Reiter and Dale, 2000; Gatt and Krahmer, 2018) . Applications of this task have become increasingly common, such as RDF-to-text , AMR-to-text (Ribeiro et al., 2019) , dialogue response generation (Dušek et al., 2018) , robo... | 0 |
A couple years ago, Peters et al. (2018, ELMo) won the NAACL Best Paper Award for creating strong performing, task-agnostic sentence representations due to large scale unsupervised pretraining. Days later, its high level of performance was surpassed by Radford et al. (2018) which boasted representations beyond a single... | 0 |
Keyphrases are simplex nouns or noun phrases (NPs) that represent the key ideas of the document. Keyphrases can serve as a representative summary of the document and also serve as high quality index terms. It is thus no surprise that keyphrases have been utilized to acquire critical information as well as to improve th... | 0 |
The relationship between a group of human languages can be characterized across several dimensions of variation (Skirgård et al., 2017) , including (1) the temporal dimension, wherein languages have diverged from a common historical ancestor as in the case of Romance languages; (2) the spatial dimension, wherein the sp... | 0 |
Remarkable progress has been achieved in many dialogue systems research disciplines, from dialogue state tracking (DST) (Dai et al., 2021; Mehri et al., 2020) to policy- (Wang et al., 2020; Lubis et al., 2020) and end-to-end modelling (Peng et al., 2021; Yang et al., 2020) . Progress is usually measured component-wise ... | 0 |
We present the results of the shared tasks of the First Conference on Statistical Machine Translation (WMT) held at ACL 2016. This conference builds on nine previous WMT workshops (Koehn and Monz, 2006; Callison-Burch et al., 2007 , 2009 , 2011 , 2012 Bojar et al., 2013 Bojar et al., , 2014 Bojar et al., , 2015 . This ... | 0 |
The task of machine reading comprehension and question answering (MRC-QA) requires the model to answer a natural language question by finding relevant information and knowledge in a given natural language context. Most MRC datasets require single-hop reasoning only, which means that the evidence necessary to answer the... | 0 |
We shall describe a general treatment of comparatives that has been implemented in the context of TELI, a question-answering system which is transportable among both domains of discourse and different types of backend retrieval systems.n Comparatives are important because of the dramatic increase in expressive power th... | 0 |
The meaning of an unknown word can often be inferred from its context. Consider the following (slightly modified) example in (Nida, 1975, p.167 (1) A bottle of tezgiiino is on the table.Everyone likes tezgiiino. Tezgiiino makes you drunk. We make tezgiiino out of corn.The contexts in which the word tezgiiino is used su... | 0 |
The work presented here is part of a larger project that has the goal of developing a suitable automated approach to verification and validation of natural language processing (NLP) systems, including structural (white-box) (Beizer, 1990) testing techniques that are suitable for language applications. In previous work ... | 0 |
Natural language communication can be used to affect change in the real world. Text adventure games, in which players must make sense of the world through text descriptions and declare actions through natural language, can provide a stepping stone toward more real-world environments where agents must communicate to und... | 0 |
Named entity recognition (NER) is one of the most fundamental natural language processing (NLP) tasks that is central to understanding unstructured, textual data. Popular approaches (Chieu and Ng, 2002; Florian et al., 2003; Ratinov and Roth, 2009; Luo et al., 2015) mainly rely on hand-crafted features and gazetteers w... | 0 |
LIMSI took part in the IWSLT 2010 evaluation for two different tasks: Talk and BTEC. The goal of the new Talk task is to translate public speeches on a variety of topics, from English to French. Since the allowed training data includes the parallel corpora distributed by the ACL 2010 Workshop on Statistical Machine Tra... | 0 |
Semantic analysis is the process of understanding the underlying meaning in text. Building a rich semantic representation is a crucial element of modern Natural Language Processing NLP. It allows us to build comprehensive text representations in order to analyse the underlying text meaning more effectively. Semantic in... | 0 |
Native Language Identification (NLI) tackles the problem of determining the native language of an author based on a text the author has written in a second language. With Tomokiyo and Jones (2001) , Jarvis et al. (2004) , and Koppel et al. (2005) as first publications on NLI, the research focus in computational linguis... | 0 |
Language evolves with time. New words appear, old words fall out of use, the meanings of some words shift. The culture changes as well as the expected audience of the printed word. There are changes in topics, in syntax, in presentation structure. Reading the natural philosophy musings of aristocratic amateurs from the... | 0 |
Manipuri language, one of the scheduled Indian languages, belonging to a Tibeto-Burman languages family (Chelliah, 1990 ) is syllabic in nature. In general, a syllable follows onset-nucleus-coda (consonant-vowel-consonant) structure, where nucleus is the core component defined by a vowel. The preceding and following co... | 0 |
Significant advances have been witnessed with neural encoder-decoder paradigm (Sutskever et al., 2014) , transformer-based architecture (Vaswani et al., 2017) and large-scale pretraining models (Devlin et al., 2019; Radford et al., 2019) in a wide array of natural language generation (NLG) tasks including machine trans... | 0 |
Quality estimation (QE) is the task of estimating the quality of given machine translations without regard to their reference translations (Blatz et al., 2004; Specia et al., 2009) . As reference translations are generally unavailable in real life, QE should help to treat output texts of machine translation (MT) system... | 0 |
Goal oriented dialogue systems help users with accomplishing tasks, like making restaurant reservations or booking flights, by interacting with them in natural language. The capability to understand user utterances and break them down into task specific semantics is a key requirement for these systems. This is accompli... | 0 |
The Duluth systems in the sense induction task of SemEval-2 (Manandhar et al., 2010) were based on SenseClusters (v1.01), a freely available open source software package which relies on the premise that words with similar meanings will occur in similar contexts (Purandare and Pedersen, 2004) . The data for the sense in... | 0 |
Morphologically inflected words are derived from a root by modifying it (e.g., by applying prefixes, suffixes and infixes) based on linguistic features (manifested as the inflection tagset). Morphological analysis involves extracting this root word and the set of features that describe the inflected form. These analyse... | 0 |
In many NLP classification applications, the classes are not symmetric and the user has some preference towards a high Precision or Recall of a particular target class. Thus, appropriate tuning of the model is often necessary, depending on the particular tolerance of the application to false positive or false negative ... | 0 |
The goal of the WMT 2012 Quality Estimation (QE) shared task (Callison-Burch et al., 2012) was to create automatic systems to judge the quality of the translations produced by a Statistical Machine Translation (SMT) system given the input text, the proposed translations and information about the models used by the SMT ... | 0 |
The LinGO Grammar Matrix (Bender et al., 2002) is a resource, rooted in the Head-driven Phrase Structure Grammar (HPSG) formalism (Pollard and Sag, 1994) , that enables linguists to create implemented precision grammars. The core of the Grammar Matrix is a collection of types and constraints expected to be cross-lingui... | 0 |
Opinion summarization aims to extract and summarize customers' opinions from reviews on products or services (Hu and Liu, 2004; Cardie et al., 2004) . With the rapid expansion of e-commerce, the number of online reviews is growing at a high speed, which makes it impractical for customers to read throughout large amount... | 0 |
The desire and need for cross-cultural communication between China and Europe, both officially and nongovernmentally, are on the increase. Especially, France is an important strategic partner of China and has deep relationships with China in many fields such as economy, science and technology, culture and education, et... | 0 |
Extending a continuous speech recognition system to a larger vocabulary and more general task domain requires more than a new dictionary and language model. The primary problem in the application of the SPHINX-II [1] system to the Wall Street Journal (WSJ) CSR task was to extend the Viterbi beam-search used in the SPHI... | 0 |
With the rapid growth of social media such as Twitter, sentiment classification towards the user generated texts has attracted increasing research interest. The objective of sentiment classification is identifying the sentiment of a text into binary polarity (Positive vs. Negative) or single-label multi-class (e.g., Ve... | 0 |
Based on Multilevel Mark Theory(1, 2) a mixture object search method is realized, it can search various objects just like searching characters, and raise Chinese text search to word level. (3, 4) UTF-8 provides a UNIX compatible transformation format of Unicode Code System (UCS), and makes UNIX systems support multi-li... | 0 |
In Machine Translation (MT), the output from multiple decoding systems can be used to create a new output which is better than any single input system, using a procedure known as system combination.Normally, the input systems are generated by varying some important aspect of the MT system, such as the alignment algorit... | 0 |
Document categorisation is a core task in information retrieval and natural language processing, whereby documents are categorised relative to a pre-defined set of labels. While the majority of research on document categorisation assumes a flat label structure, in practice in large-scale document categorisation tasks, ... | 0 |
Children's spelling has long been considered to be rooted in phonetics; being influenced by pronunciation and speech sound similarity (Read, 2018) . Misspellings are often 'creative' in that they reflect the phonetic judgements of the child and stem from their efforts to use letters to represent the sounds of their spo... | 0 |
During the last decade, social media have become extremely popular, on which billions of usergenerated contents are posted every day. Many users have been writing about their thoughts and lives on the go. The massive unstructured data from social media provides valuable information for a variety of applications such as... | 0 |
Recently, several studies have suggested that document context is required for the reliable human evaluation of machine-translated documents Laubli et al., 2020) . With the improved performance of neural machine translation systems (NMT) over the past years, this is particularly important when assessing the potential f... | 0 |
Although Irish is a low-resourced language, the Irish Government's 20 Year Strategy for the Irish Language, which prioritises the "promotion and protection" of the language (Government of Ireland, 2010) , has brought about investment in the creation of digital language tools and resources. Linguistic resources, such as... | 0 |
In statistical machine translation (SMT) (Brown et al., 1993) , the translation model is automatically learned form parallel corpora in an unsupervised way. The translation model contains translation pairs with their features scores. SMT suffers from the accuracy problem that the translation model may be inaccurate, me... | 0 |
Classifier-based dependency parsers (Yamada and Matsumoto, 2003; Nivre and Scholz, 2004) learn from an annotated corpus how to select an appropriate sequence of Shift/Reduce actions to construct the dependency tree for a sentence. Learning is based on techniques such as SVM (Vapnik 1998) or Memory Based Learning (Daele... | 0 |
Forms are a ubiquitous document format. Numerous forms are used in finance, insurance and medical industry every day. Although forms vary a lot, we consider it as a collection of key-value pairs and all these pairs establish a hierarchical structure within the page. Our work in this paper focuses on utilizing multimoda... | 0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.