text stringlengths 4 222k | label int64 0 4 |
|---|---|
In recent years, with the advent of large training corpora and pretrain technology, chatbot models have evolved considerably in open domain (Bao et al., 2020; Roller et al., 2021) . Current chatbots have achieved surprising results in generating fluent, engaging, informative responses, but still occasionally generate r... | 0 |
The Expedition project is devoted to fast "ramp-up" of machine translation systems from less studied, so-called "low-density" languages into English. One of the components that must be acquired and built during this process is a morphological analyzer for the source low-density language. Since we expect that the source... | 0 |
Semantic analyses often go beyond treestructured representations, assigning multiple semantic heads to nodes, some semantic formalisms even tolerating directed cycles. 1 At the same time, syntactic parsing is a mature field with efficient, highly optimised decoding and learning algorithms for tree-structured representa... | 0 |
Parmi l'ensemble des outils d'aide à la traduction, le système de mémoire de traduction (SMT) est certainement l'outil le plus populaire auprès des traducteurs professionnels. Comme l'explique (Planas, 2000) , ce succès est dû à deux types de redondances que le traducteur rencontre fréquemment dans son activité et qui ... | 0 |
The University of Edinburgh participated in the WMT19 Shared Task on News Translation in six language directions: English-Gujarati (EN↔GU), English-Chinese (EN↔ZH), German-English (DE→EN) and English-Czech (EN→CS). All our systems are neural machine translation (NMT) systems trained in constrained data conditions with ... | 0 |
People can discriminate between typical (e.g., A cop arrested a thief ) and atypical events (e.g., A thief arrested a cop) and exploit this ability in online sentence processing to anticipate the upcoming linguistic input. Brains have been claimed to be "prediction machines" (Clark, 2013) and psycholinguistic research ... | 0 |
Twitter has become one of the most widely used social media platforms, with users (as of March 2013) posting approximately 400 million tweets per day (Wickre, 2013) . This public data serves as a potential source for a multitude of information needs, but the sheer volume of tweets is a bottleneck in identifying relevan... | 0 |
Sentiment analysis of English texts has become a large and active research area, with many commercial applications, but the barrier of language limits the ability to assess the sentiment of most of the world's population.Although several well-regarded sentiment lexicons are available in English (Esuli and Sebastiani, 2... | 0 |
Machine learning models using deep neural architectures have seen tremendous performance improvements over the last few years. The advent of models such as LSTMs (Hochreiter and Schmidhuber, 1997) and, more recently, attention-based models such as Transformers (Vaswani et al., 2017) have allowed some language technolog... | 0 |
In recent times, we have seen how Internet has revolutionized the field of education through Massive Open Online Courses (MOOCs). Universities are incorporating MOOCs as a part of their regular coursework. Since most of these courses are in English, the students are expected to know the language before they are admitte... | 0 |
Manually annotated corpora and treebanks are the primary tools that we have for developing and evaluating models and theories for natural language processing. Given their importance for testing our hypotheses, it is imperative that they are of the best quality possible. However, manual annotation is tedious and error-p... | 0 |
The increasing amount of documents in electronic form makes imperative the need for document content classification and semantic labelling. Keyphrase extraction contributes to this goal by the identification of important and discriminative concepts expressed as keyphrases. Keyphrases as reduced document content represe... | 0 |
Over the last years, corpus based approaches have gained significant importance in the field of natural language processing (NLP). Large corpora for many different languages are currently being collected all over the world, like In order to use this amount of data for training and testing purposes of NLP systems, corpo... | 0 |
Great strides have been made in Speech-to-Speech (S2S) translation systems that facilitate cross-lingual spoken communication [1] [2] [3] . While these systems [3] [4] [5] already fulfill an important role, their widespread adoption requires broad domain coverage and unrestricted dialog capability. To achieve this, S2S... | 0 |
Natural language understanding seeks to create models that read and comprehend text. A common strategy for assessing the language understanding capabilities of comprehension models is to demonstrate that they can answer questions about documents they read, akin to how reading comprehension is tested in children when th... | 0 |
Many existing approaches to text generation rely on recurrent neural networks trained using likelihood on sequences of words or characters. However, such models often fail to capture overall structure and coherency in multi-sentence or longform text Holtzman et al., 2018) . To rectify this, prior work has proposed loss... | 0 |
With the translation industry exponentially growing, more hope is vested in the use of machine translation (MT) to increase translators' productivity (Rinsche and Portera-Zanotti, 2009) . Though post-editing MT has proven to increase productivity and even quality for certain text types (Tatsumi, 2010) , research on the... | 0 |
Emojis have become crucial components of written language. Emojis were initially designed to express emotions or feelings, e.g., for a smiley face, and they have grown to be a large family of over 2,000 icons over the years which can express not only emotions but a wide range of objects or actions, e.g., for a gift and... | 0 |
Cross-lingual text representations have become extremely popular in NLP, since they promise universal text processing in multiple human languages with labeled training data only in a single one. They go back at least to the work of Klementiev et al. (2012) , and have seen an exploding number of contributions in recent ... | 0 |
Quality Estimation (QE) for Machine Translation (MT) (Blatz et al., 2004; Quirk, 2004; Specia et al., 2009) aims at providing quality scores or labels to MT output when translation references are not available. Sentence-level QE is usually conducted using human produced direct assessments (DA) (Graham et al., 2013) or ... | 0 |
An event is something that occurs in a certain place at a certain time (Pustejovsky et al., 2003) . Understanding events plays a major role in various natural language processing tasks such as information extraction (Humphreys et al., 1997) , question answering (Narayanan and Harabagiu, 2004) , textual entailment (Hagh... | 0 |
Paraphrase generation, namely rewriting a sentence using different words and/or syntax while preserving its meaning (Bhagat and Hovy, 2013) , is an important technique in natural language processing, that has been widely used in various downstream tasks including question answering (Fader et al., 2014a; McCann et al., ... | 0 |
Prosody refers to the suprasegmental features of natural speech, such as rhythm and intonation, since it normally extends over more than one phoneme segment. Speakers use prosody to convey paralinguistic information such as emphasis, intention, attitude, and emotion. Humans listening to speech with natural prosody are ... | 0 |
Quality Estimation (QE) aims to predict the quality of the output of Machine Translation (MT) systems when no gold-standard translations are available. It can make MT useful in real-world applications by informing end-users on the translation quality. We focus on sentence-level QE, usually formulated as a regression ta... | 0 |
Grammatical and linguistic acceptability is an extensive area of research that has been studied for generations by theoretical linguists (e.g. Chomsky, 1957) , and lately by cognitive and compu-1 SwedishGlue (Swe. SuperLim) is a collection of datasets for training and/or evaluating language models for a range of Natura... | 0 |
Tables convey important information in a concise manner. This is true in many domains, scientific documents being one of them. Truth verification tasks in past(e.g. SemEval-2019 Fact Checking Task) have focused on written text without considering the tables. The current shared task (Wang et al., 2021) focuses on tables... | 0 |
The Statistical machine translation (SMT) systems are considered as one of the most popular approaches to machine translation (MT). However, SMT can suffer from grammatically incorrect output with erroneous syntactic and semantic structure for the language pair on which it is being applied. It is observed that the gram... | 0 |
Spoken language translation (SLT) connects automatic speech recognition (ASR) and machine translation (MT) by translating recognized spoken language into a target language. In general, the speech translation process is divided into two separate parts. First, an ASR system provides an automatic transcription of spoken w... | 0 |
The naive Bayes classifier has been one of the core frameworks in the information retrieval research for many years. Recently, naive Bayes is emerged as a research topic itself because it sometimes achieves good performances on various tasks, compared to more complex learning algorithms, in spite of the wrong independe... | 0 |
Grammatical formalisms such as HPSG [Pollard and Sag, 1987] [Pollard and Sag, 1992] and LFG [Kaplan and Bresnan, 1982] employ feature descriptions [Kasper and Rounds, 1986] [Smolka, 1992] as the primary means for stating linguistic theories. However the descriptive machinery employed by these formalisms easily exceed t... | 0 |
Words, syntactic groups, clauses, sentences, paragraphs, etc. usually form the basis of the analysis and processing of natural language text. However, texts in electronic form are just sequences of characters, including letters of the alphabet, numbers, punctuation, special symbols, whitespace, etc. The identification ... | 0 |
Reordering remains one of the greatest challenges in Statistical Machine Translation (SMT) research as the key contextual information may span across multiple translation units. 1 Unfortunately, previous approaches fall short in capturing such cross-unit contextual information that could be critical in reordering. For ... | 0 |
Minimum Bayes Risk (MBR) is a theoreticallyelegant decision rule that has been used for singlesystem decoding and system combination in machine translation (MT). MBR arose in Bayes decision theory (Duda et al., 2000) and has since been applied to speech recognition (Goel and Byrne, 2000) and machine translation (Kumar ... | 0 |
Traditional written corpora for linguistics research are created primarily from printed text, such as newspaper articles and books. With the growth of the World Wide Web as an information resource, it is increasingly being used as training data in Natural Language Processing (NLP) tasks.There are many advantages to cre... | 0 |
In recent years, advances in the field of Natural Language Processing (NLP) have revolutionized the way machines are used to interpret humanwritten text. With the rapid accumulation of publicly available documents, from newspaper articles to social media posts, machine learning methods designed to automate data analysi... | 0 |
Rooted in Structural Linguistics (de Saussure, 1966; Harris, 1951) , Distributional Semantic Models (DSMs, see e.g. (Baroni and Lenci, 2010) ) characterize the meaning of lexical units by the contexts they appear in, cf. (Wittgenstein, 1963; Firth, 1957) . Using the duality of form and contexts, forms can be compared a... | 0 |
Deep neural networks have achieved remarkable successes in natural language processing recently. Although neural models have demonstrated performance superior to humans on some tasks, e.g. reading comprehension (Rajpurkar et al., 2016; Devlin et al., 2019; Lan et al.) , it still lacks the ability of discrete reasoning,... | 0 |
The success of statistical approaches to Machine Translation (MT) can be attributed to the IBM models (Brown et al., 1993) that characterize wordlevel alignments in parallel corpora. Parameters of these alignment models are learnt in an unsupervised manner using the EM algorithm over sentence-level aligned parallel cor... | 0 |
Medical dialogue systems, which have gained increasing attention, aim to communicate with patients to enquire about diseases beyond their selfreported and make an automatic diagnosis (Wei et al., 2018; Lin et al., 2019) . It has the potential to substantially automate the diagnostic process while also lowering the cost... | 0 |
The purpose of an information retrieval (IR) system is to retrieve the documents relevant to user's information need expressed in the form of a query. Many information needs are event-oriented, while at the same time there exists an abundance of event-centered texts (e.g., breaking news, police reports) that could sati... | 0 |
Many algorithms in speech and language processing can be viewed as instances of dynamic programming (DP) (Bellman, 1957) . The basic idea of DP is to solve a bigger problem by divide-and-conquer, but also reuses the solutions of overlapping subproblems to avoid recalculation. The simplest such example is a Fibonacci se... | 0 |
Bilingual data (including bilingual sentences and bilingual terms) are critical resources for building many applications, such as machine translation (Brown, 1993) and cross language information retrieval (Nie et al., 1999) . However, most existing bilingual data sets are (i) not adequate for their intended uses, (ii) ... | 0 |
Vowel distinction exists due to their being different vowel inventories and phonetic features across languages. There are new and similar vowels when comparing two vowel systems of languages. Similar vowels represent the vowels sharing certain phonetic features and phonology status within two vowel systems. While non-n... | 0 |
Event trigger extraction, as defined the Automatic Content Extraction multilingual evaluation benchmark (ACE2005) (Walker, 2006) , is a subtask of event extraction which requires systems to detect and label the lexical instantiation of an event, known as a trigger. As an example, in the sentence "John traveled to NYC f... | 0 |
The task of relation extraction (RE) deals with identifying whether any pre-defined semantic relation holds between a pair of entity mentions in the given sentence. Pure relation extraction techniques (Zhou et al., 2005; Jiang and Zhai, 2007; Bunescu and Mooney, 2005; Qian et al., 2008) assume that for a sentence, gold... | 0 |
Many natural language phenomena are inherently not context-free, or call for structural descriptions that cannot be produced by a context-free grammar (Chomsky, 1957; Shieber, 1985; Savitch et al., 1987) . Examples are extraposition, cross-serial dependencies and WH-inversion. However, relaxing the context-freeness ass... | 0 |
Bilingual Word Embeddings are useful for crosslingual tasks such as cross-lingual transfer learning or machine translation. Mapping based BWE approaches rely only on a cheap bilingual signal, in the form of a seed lexicon, and monolingual data to train monolingual word embeddings (MWEs) for each language, which makes t... | 0 |
Intelligent tutoring systems help students improve learning compared to reading textbooks, though not quite as much as human tutors (Anderson et al., 1995) . The specific properties of human-human dialogue that help students learn are still being studied, but the proposed features important for learning include allowin... | 0 |
A common sequence-labeling task in natural language processing involves assigning a part-ofspeech (POS) tag to each word in the input text. Previous authors have used numerous HMM-based models (Banko and Moore, 2004; Collins, 2002; Lee et al., 2000; Thede and Harper, 1999) and other types of networks including maximum ... | 0 |
Commercial search engines use query associations in a variety of ways, including the recommendation of related queries in Bing, 'something different' in Google, and 'also try' and related concepts in Yahoo. Mining techniques to extract such query associations generally fall into four categories: (a) clustering queries ... | 0 |
The main contribution of this paper is the presentation of a conceptualized and implemented workflow for the study of relations between entities mentioned in text. The workflow has been realized for multiple, diverse but structurally similar research questions from Humanities and Social Sciences, although this paper fo... | 0 |
For a long time, speech has been the only modality for input and output in telephone-based information systems. Speech is often considered to be the most natural form of input for such systems, since people have always used speech as the primary means of communication. Moreover, to use a speech-only system a simple tel... | 0 |
The goal of the Recognizing Textual Entailment (RTE) task is, given a pair of sentences, to determine whether a Hypothesis sentence can be inferred from a Text sentence. The majority of work in RTE is focused on finding a generic solution to the task. That is, creating a system that uses the same algorithm to return a ... | 0 |
Polish, one of the West-Slavic languages [1] , due to its complex inflection and free word order, forms a challenge for statistical machine translation (SMT). Polish grammar is quite complex: seven cases, three genders, animate and inanimate nouns, adjectives agreed with nouns in terms of gender, case and number and a ... | 0 |
New Dimensions in Testimonies (NDT) is a dialogue system that allows for two-way communication with a person who is not available for conversation in real time: a large set of statements is prepared in advance, and users access these statements through natural conversation that mimics face-to-face interaction (Artstein... | 0 |
Evaluating text difficulty, or text readability, is an important topic in natural language processing and applied linguistics (Zamanian and Heydari, 2012; Pitler and Nenkova, 2008; Fulcher, 1997) . A key challenge of text difficulty evaluation is that linguistic difficulty arises from both vocabulary and grammar (Richa... | 0 |
Text summarization is a major NLP task, whose aim is "to present the main ideas in a document in less space" (Radev et al., 2002) . A specific type of summarization is the extractive summarization, which consists of selecting a subset of the original sentences of a document for verbatim inclusion in the summary; in con... | 0 |
The Arabic language is characterized by diglossia (Ferguson, 1959) : two linguistic variants live side by side: a standard written form and a large variety of spoken dialects. While dialects differ from one region to another, the written variety, called Modern Standard Arabic (MSA), is generally the same. MSA, the offi... | 0 |
Recently, the capability of large-scale pre-trained models has been verified in open-domain dialogue generation, including Meena (Adiwardana et al., 2020) , Blender (Roller et al., 2021) , and PLATO-2 (Bao et al., 2020) . Without introducing explicit knowledge in learning process, substantive knowledge is implicitly em... | 0 |
Dialogue Acts (DAs) are the functions of utterances in dialogue-based interaction (Austin, 1975) . A DA represents the meaning of an utterance at the level of illocutionary force, and hence, constitutes the basic unit of linguistic communication (Searle, 1969) . DA classification is an important task in Natural Languag... | 0 |
Computational models of language understanding must recognize narrative structure because many types of natural language texts are narratively structured, e.g. news, reviews, film scripts, conversations, and personal blogs (Polanyi, 1989; Jurafsky et al., 2014; Bell, 2005; Gordon et al., 2011a) . Human understanding of... | 0 |
We present an algorithm for identifying noun phrase antecedents of personal pronouns, demonstrative pronouns, reflexive pronouns, and omitted pronouns (zero pronouns) in Spanish. The algorithm identifies both intrasentential and intersentential antecedents and is applied to the syntactic analysis generated by the slot ... | 0 |
La génération textuelle est une tâche centrale pour l'interaction entre un système intelligent et ses utilisateurs (réponse d'un agent conversationnel, résumé de texte, génération d'article...). Lors de cette interaction, il est désirable de contrôler les générations afin qu'elles respectent des contraintes imposées pa... | 0 |
Previous work on Chinese Semantic Role Labeling (SRL) mainly focused on how to implement SRL methods which are successful on English. Similar to English, parsing is a standard pre-processing for Chinese SRL. Many features are extracted to represent constituents in the input parses (Sun and Jurafsky, 2004; Xue, 2008; Di... | 0 |
In the biomedical domain, the vast amount of data and the great variety of induced features are two major bottlenecks for further natural language processing on the biomedical literature. In this paper, we investigate the biomedical named entity recognition (NER) problem. This problem is particularly important because ... | 0 |
In modern business, contact centers are becoming more and more important for improving customer satisfaction. Such contact centers typically have quality analysts who mine calls to gain insight into how to improve business productivity (Takeuchi et al., 2007; Subramaniam et al., 2009) . To enable them to handle the mas... | 0 |
Persuasion is a primary goal of argumentation (O'Keefe, 2006) . It is often carried out in the form of a debate or discussion, where debaters argue to persuade others to take certain stances on controversial topics. Several studies have examined persuasiveness in debates by probing the main factors for establishing per... | 0 |
Many existing dialogue systems adopt a two-phase approach to satisfying a user's request for information: query construction followed by solution construction and presentation, with the former concerned with the acquisition of the preferences and restrictions in a user's information needs, and the latter concerned with... | 0 |
It has usually been assumed that the semantics of temporal expressions is directly related to the linear dimensional concaption of time familiar from high-school physics -that is, to a model based on the number-line. However, there are good reasons for suspecting that such a conception is not the one that our linguisti... | 0 |
Obtaining the definition is the first step toward understanding a new terminology. The lack of precise terminology definition poses great challenges in scientific communication and collaboration (Oke, 2006; Cimino et al., 1994) , which further hinders new discovery. This problem becomes even more severe in emerging res... | 0 |
of all public-facing documents, while sociocultural pressure has also influenced private businesses to invest in translation services. But the mounting demand for translation services presents challenges as well as opportunities for Welsh Linguistic Service Providers (hereafter LSPs). LSPs need to balance expenditure (... | 0 |
In the context of goal-oriented dialogue systems, intent classification (IC) is the process of classifying a user's utterance into an intent, such as Book-Flight or AddToPlaylist, referring to the user's goal. While slot filling (SF) is the process of identifying and classifying certain tokens in the utterance into the... | 0 |
Vector space models and distributional information have been a steadily increasing, integral part of lexical semantic research over the past 20 years. On the one hand, vector space models (see Turney and Pantel (2010) and Erk (2012) for two recent surveys) have been exploited in psycholinguistic (Lund and Burgess, 1996... | 0 |
Sarcasm detection is the computational task of predicting sarcasm in text. Past approaches in sarcasm detection rely on designing classifiers with specific features (to capture sentiment changes or incorporate context about the author, environment, etc.) Wallace et al., 2014; Rajadesingan et al., 2015; Bamman and Smith... | 0 |
Distributed representation of words (Bengio et al., 2003; Mikolov et al., 2013a; Pennington et al., 2014; Bojanowski et al., 2017) and sentences (Kiros et al., 2015; Conneau et al., 2017; Reimers and Gurevych, 2019; Gao et al., 2021) have shown to be extremely useful in transfer learning to many NLP tasks. Therefore, i... | 0 |
Extracting knowledge from unstructured text has been a long-standing goal of NLP and AI. The advent of the World Wide Web further increases its importance and urgency by making available an astronomical number of online documents containing virtually unlimited amount of knowledge (Craven et al., 1999) . A salient examp... | 0 |
Discourse structures for texts represent relational semantic structures that convey causal, topical, argumentative relations inter alia or more generally coherence relations. Following (Muller et al., 2012; Li et al., 2014; Morey et al., 2018) , we represent them as dependency structures or graphs containing a set of n... | 0 |
There is a long linguistic tradition of frame and role annotation for verbal predications, rooted in verb sense classifications on the one hand (e.g. Levin 1993), and the concept of semantic roles (also called thematic or case roles, Fillmore 1968) on the other. In a frame-based framework, verb categories and semantic ... | 0 |
Linguistic analyses have often been computationally performed around the static notion of words or word categorization methods (e.g. LIWC) where the context of words is not taken into account. Previous work on politeness (Danescu-Niculescu-Mizil et al., 2013) and gender (Bamman et al., 2014) have focused on comparing t... | 0 |
Speech-to-text translation (ST) has been traditionally approached with cascade architectures consisting of a pipeline of two sub-components (Stentiford and Steer, 1988; Waibel et al., 1991) : an automatic speech recognition (ASR), which transforms the audio input into a textual representation, and a machine translation... | 0 |
Large pre-trained language models (LMs) are continuously pushing the state of the art across various NLP tasks. The established procedure performs self-supervised pre-training on a large text corpus and subsequently fine-tunes the model on a specific target task (Devlin et al., 2019; Liu et al., 2019b) . The same proce... | 0 |
Named entity recognition (NER) is a fundamental task in information extraction, and the ability to detect mentions of domain-relevant entities such as chemicals and proteins is required for the analysis of texts in specialized domains such as biomedicine. Although a wealth of manually annotated corpora and dedicated NE... | 0 |
Currently a major part of cutting-edge research in MT revolves around the statistical machine translation (SMT) paradigm. SMT has been inspired by the use of statistical methods to create language models for a number of applications including speech recognition. A number of different translation models of increasing co... | 0 |
The functionality of systems that extract information from texts can be specified quite simply: the input is a stream of texts and the output is some representation of the information to be extracted. In the message understanding research promoted by ARPA through its Human Language Technology initiative, the form of th... | 0 |
According to the World Health Organization (WHO), 20% of children and adolescents in the world have mental disorders or problems (WHO, 2014) . Suicide ranks as the second leading cause of death in the 15-29 years old group and every 40 seconds a person dies by suicide in the world. The WHO pointed early identification ... | 0 |
Figure 1: Examples of disambiguating information provided by images for the prepositional phrase attachment of the sentence Mary eats spaghetti with a friend (Gokcen et al., 2018) .that the proposed models achieve state-of-the-art results on multilingual induction datasets, even without help from linguistic knowledge o... | 0 |
Dialogue systems based on deep neural networks (DNNs) have been widely studied. Although these dialogue systems can generate fluent responses, they often generate dull responses such as "yes, that's right" and lack engagingness as a conversation partner (Jiang and de Rijke, 2018) . To develop an engaging dialogue syste... | 0 |
Paraphrase generation is an important and challenging task in the field of Natural Language Processing (NLP), which can be applied in a variety of applications such as information retrieval (Yan et al., 2016) , question answering (Fader et al., 2014; Yin et al., 2015) , machine translation (Cho et al., 2014) , and so o... | 0 |
Text, no matter the length, can potentially convey an emotional meaning. As the availability of digitized documents has increased over the past decade, so the ability and need to classify this data by its affective content has increased. This in turn has generated a large amount of interest in the field of Sentiment An... | 0 |
As China plays a more and more important role of the world, learning Chinese as a foreign language is becoming a growing trend, which brings opportunities as well as challenges. Due to the variousity of grammar and the flexibility of expression, Chinese Grammatical Error Dignosis(CGED) poses a serious challenge to both... | 0 |
Text classification is one of the most fundamental tasks in natural language processing (NLP). In real-world scenarios, labeling massive texts is timeconsuming and expensive, especially in some specific areas that need domain experts to participate. Weakly-supervised text classification (WTC) has received much attentio... | 0 |
Many of the world's languages have a rich morphology, i.e., make use of surface variations of lemmata in order to express certain properties, like the tense or mood of a verb. This makes a variety of natural language processing tasks more challenging, as it increases the number of words in a language drastically; a pro... | 0 |
Chinese sentences arc cx)mposed with string of characters without blanks to mark words. However the basic unit for sentence parsing and understanding is word. Therefore the first step of processing Chinese sentences is to identify the words( i.e. segment the character strings of the sentences into word strings).Most of... | 0 |
Aligned corpora have proved very useful in many tasks, including statistical machine translation, bilingual lexicography (Daille, Gaussier and Lange 1993), and word sense disambiguation (Gale, Church and Yarowsky 1992; Chen, Ker, Sheng, and Chang 1997). Several The statistical approach to machine translation (SMT) can ... | 0 |
With the growing popularity of Twitter, sentiment analysis of tweets has drawn the attention of several researchers from both academia and industry in recent times. Unlike other regular texts, sentiment analysis on Twitter text poses plenty of challenges because of various characteristics such as (i) under-specificity ... | 0 |
Syntactically annotated treebanks are a great resource of linguistic information that is available hardly or not at all in flat text corpora. Retrieving this information requires specialized tools. Some of the best-known tools for querying treebanks include TigerSEARCH (Lezius, 2002) , TGrep2 (Rohde, 2001) , MonaSearch... | 0 |
The popularity of speech synthesis as a topic in natural language processing has significantly increased after the publication of results by DeepMind (van den Oord et al., 2016) , Baidu (Arik et al., 2017) and Google (Wang et al,, 2017; Shen et al., 2018) , demonstrating the ability to create natural sounding speech wi... | 0 |
Humor is an important ingredient of human communication, and every automatic system aiming at emulating human intelligence will eventually have to develop capabilities to recognize and generate humorous content. In the artificial intelligence community, research on humor has been progressing slowly but steadily. As an ... | 0 |
Consider the following example from the Trains corpus (Heeman and Allen 1995) .Example 1 (d93-13.3 utt63) um it'll be there it'll get to Dansville at three a.m. and then you wanna do you take tho-want to take those back to Elmira so engine E two with three boxcars will be back in Elmira at six a.m. is that what you wan... | 0 |
Improvements in data-driven parsing approaches, coupled with the development of treebanks that serve as training data, have resulted in accurate parsers for several languages. However, portability across domains remains a challenge: parsers trained using a treebank for a specific domain generally perform comparatively ... | 0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.