text stringlengths 4 222k | label int64 0 4 |
|---|---|
Word is the fundamental unit in natural language understanding. Since people do not retain the boundary information between words in practical use, Chinese Word Segmentation (CWS) is the very first step in Chinese information processing. A considerable amount of research has shown that using character sequence labeling... | 0 |
This paper provides an overview of the PROMT submissions for the WMT 2020 Shared News Translation Task. This year we participate with neural MT systems for the third time. We participate in four language pairs and six directions. We describe our data preparation pipelines, models training setups and present the results... | 0 |
High-quality translation dictionaries are an indispensable resource in both language technology and language learning applications. For instance, rule-based translation systems (Forcada et al., 2011; Mayor et al., 2011; Popel andŽabokrtský, 2010 ) rely on high-quality dictionaries. Unlike statistical translation system... | 0 |
In the last couple of years our group at DFKI in Saarbrücken has been involved in a number of projects aiming at interfacing different devices in an intelligent way. The main goal of these projects has been to build functioning robust systems with which it is natural to communicate (not only for some few examples phras... | 0 |
Social media platforms provide a versatile medium for people to communicate with each other, share ideas and express opinions. These user-driven platforms have a challenge when it comes to controlling the content being fed into them. People have different intentions, while some might use these platforms for their inten... | 0 |
Dependency parsing has provided new methods and resources for natural language technology tasks in recent years. Dependency treebanks are now available for many languages and parsed data is used for improving machine translation, search engines and other NLP applications. While data resources are relatively common for ... | 0 |
This paper highlights the development of the Dutch Parallel Corpus (DPC): a high quality sentence-aligned parallel corpus of 10 million words for the language pairs Dutch-English and Dutch-French.Since the development of a parallel corpus is timeconsuming and costly, the DPC project aims at the creation of a multifunct... | 0 |
As an essential aspect of artificial intelligence, dialogue systems have attracted extensive attention in recent studies (Vinyals and Le, 2015; Serban et al., 2016) . Researchers have paid great efforts to understand conversations between two participants, either single-turn (Li et al., 2016a; Shang et al., 2015; Vinya... | 0 |
User generated reviews on products are expanding rapidly with the emergence and advancement of ecommerce. These reviews are valuable to business organizations for improving their products and to individual consumers for making informed decisions. Unfortunately, reading though all the product reviews is hard, especially... | 0 |
User-friendly information and communications technologies and easily available access to the Internet were critical in development of Social Web, a socio-technical phenomenon that enables people to connect, support and learn from each other (Ho et al, 2014) . The world-wide social media helped to create a digital resou... | 0 |
Contemporary task-oriented dialog systems rely on natural language understanding (NLU) models to classify a user's utterance into an intent class (e.g., "what is my savings account balance" → get balance) as well as to extract important entities from the utterance (e.g., "savings account" → BANK ACCOUNT). While the cho... | 0 |
Of the software packages that provide support for working with language corpora, the NITE XML Toolkit (NXT) is arguably the most mature, offering to combine multiple audio and video signals with crossing structures of linguistic annotation. It is currently in use on a range of corpora. Current users both create annotat... | 0 |
Clinical concept extraction involves the identification of sequences of terms which express meaningful concepts in a clinical setting. The identification of such concepts is important for enabling secondary usage of reports of patient treatments and interventions, e.g., in the context of cancer monitoring and reporting... | 0 |
Web-forums are rich sources of user-generated content on the web. The increasing popularity of technical forums have motivated major IT companies like Symantec to create and support forums around their products and services. For individual users or larger customers, such forums provide an easy source of information and... | 0 |
Un segment d'information évolutive (SEDIS-) est une portion de texte de longueur variable contenant de l'information dont la particularité est d'être susceptible d'évoluer dans le temps. Cette notion d'évolutivité de l'information a pris naissance dans un contexte industriel particulier, l'édition, et plus précisément ... | 0 |
Ethical challenges related to dialogue systems and conversational agents raise novel research questions, such as learning from biased data sets (Henderson et al., 2018) , and how to handle verbal abuse from the user's side (Cercas Curry and Rieser, 2018; Angeli and Brahnam, 2008; Angeli and Carpenter, 2006; Brahnam, 20... | 0 |
Research on training parsers from annotated data has for the most part focused on models and training algorithms for phrase structure parsing. The best phrase-structure parsing models represent generatively the joint probability P (x, y) of sentence x having the structure y (Collins, 1999; Charniak, 2000) . Generative ... | 0 |
Understanding temporal relationships between events in a passage is essential for natural language understanding Dong et al., 2019) . Temporal reading comprehension (TRC) (Ning et al., 2020 ) is a natural way to study temporal relations since natural language questions are flexible to capture divergent temporal relatio... | 0 |
Oculomotor control in reading has been a wellestablished domain in eye tracking research. From the perspective of perceptual and cognitive mechanisms that drive eye movement control, the characteristics of the visual stimuli is better controlled in reading research than visual scene stimuli. Several computational model... | 0 |
present a framework for translation based on the description and correspondence concepts of Lexical-Functional Grammar (Kaplan and Bresnan, 1982) . LFG formulates the syntactic dependencies and generalizations of natural languages in terms of the properties of formal structures of different types: ordinary phrase-struc... | 0 |
Natural language generation (NLG) is the task of generating natural language from a machineinterpretable meaning representation (MR). Traditionally, NLG systems tend to be rule-based and require domain-specific language resources (i.e. sentence plans, aggregation rules, ordering information, etc.) to be manually author... | 0 |
This year, we participated in the restricted translation task at WAT 2021 (Nakazawa et al., 2021) , in which we were asked to control a model so that the translation output would contain specified terms. Although the recent neural machine translation (NMT) model achieves excellent performance, controlling its output is... | 0 |
There has been no research conducted for measuring similarity between short sentences written in Sinhala, an official language of Sri Lanka, which is currently used by a population of over 16 million. Several unsupervised techniques are used for short sentence similarity calculations. These unsupervised approaches can ... | 0 |
Electronic Health Record (EHR) systems have become invaluable repositories of patient information, but their poor design and inadequate functionality make it difficult for physicians to assimilate the vast amounts of data, reducing physician productivity and negatively impacting patient care [1] [2] . Advanced clinical... | 0 |
Mental lexicon is the representation of the words in the human mind and their associations that help fast retrieval and comprehension (Aitchison, 1987) . Words are known to be associated with each other in terms of, orthography, phonology, morphology and semantics. However, the precise nature of these relations is unkn... | 0 |
Up to now, most work on statistical parsing has been focussed on English, a language with a configurational word order and little morphology. The inherent properties of morphologically rich languages include a higher variability in structure due to less rigid word order constraints, thus leading to greater attachment a... | 0 |
Inspired by the learning behavior of humans, Curriculum Learning (CL) for neural network training starts from a basic idea of "starting small", namely better to start from easier aspects of a task and then progress towards aspects with increasing level of difficulty (Elman, 1993) . Bengio et al. (2009) achieves signifi... | 0 |
With the advent of smart conversational agents such as Amazon Alexa, Google Assistant, etc., dialogue systems are becoming ubiquitous. In the context of enterprises, the majority of these systems target task oriented dialogues with the user trying to achieve a goal, e.g. booking flight tickets or ordering food. Natural... | 0 |
Previous studies have shown that treebanks can be helpful when constructing grammars. The most well-known example is PCFG-based statistical parsing (Charniak and Johnson, 2005) , where a PCFG is induced from, for instance, the Penn Treebank. The underlying statistical techniques have been refined in the last decade, an... | 0 |
The language use of speakers is related to variables such as the speaker's gender and age (Eckert, 1997; Eckert and McConnell-Ginet, 2013) . Systems that can automatically predict such variables have been receiving increasing attention. They enable more fine-grained analyses of trends by profiling the involved users. T... | 0 |
Visually grounded metaphors are metaphors that function on the visual similarity between two objects. Humans use visually grounded metaphors to describe scenes or objects in a vivid way: we could call a large person an elephant, describe somebody's blue eyes as a clear summer sky; and so forth. The basic idea is that u... | 0 |
Recently, statistical machine translation (SMT) models, especially the phrase based translation models [1] , have been widely used and have achieved great improvements. However, there are still some hard problems. One of them is how to translate OOV words. Among all OOV words, the numerical and time expressions (we gen... | 0 |
Automatic grading systems have been in practice in the educational domain for many years now, but primarily for recognition questions where students have to choose the correct answer from given options such as multiple choice questions (MCQs). Prior research has shown that such recognition questions are deficient as th... | 0 |
The International Workshop on Spoken Language Translation (IWSLT) is an evaluation campaign organized by the Consortium for Speech Translation Advanced Research (C-STAR) 2 , that provides a common framework to compare and improve current state-of-the-art speech-to-speech translation technologies. Previous IWSLT worksho... | 0 |
Extracting information about entities and their relationships from unstructured text is an important problem in NLP. Conventional datasets and methods for information extraction (IE) focus on within-sentence relations from general Newswire text (Zhang et al., 2017) . However, recent work started studying the developmen... | 0 |
Information extraction (IE) aims at generating structured data from free text in order to populate Knowledge Bases (KBs). Hence, one is given an incomplete KB composed of a set of triples of the form (h, r , t); h is the left-hand side entity (or head), t the right-hand side entity (or tail) and r the relationship link... | 0 |
In multimodal machine translation, a target sentence is translated from a source sentence together with related nonlinguistic information such as visual information. Recently, neural machine translation (NMT) has superseded traditional statistical machine translation owing to the introduction c ⃝ 2019 The authors. This... | 0 |
The Unaccusative Hypothesis (UH), initially proposed by Perlmutter (1978) within Relational Grammar and later elaborated on by Burzio (1986) within Government and Binding Theory, classifies intransitive verbs into unergatives and unaccusatives based on their different syntactic and semantic properties. However, myriad ... | 0 |
Given the limitations of computational and human resources, state-of-the-art statistical taggers mostly use context information on tag bigrams, for smaller training corpora, or trigrams, for larger training corpora. Some also use lexical emission probabilities above unigrams, although with a rather limited context view... | 0 |
One problem with popular word embedding methods such as word2vec (Mikolov et al., 2013) and GloVe (Pennington et al., 2014) is that they assign polysemic or homonymic words the same vector representation, i.e., words that share the same spelling but have different meanings obtain the same representation. For example, t... | 0 |
Understanding math problems via automated methods is a desired machine capacity for artificial intelligence assisted learning. Such a capacity is the key to the success of a variety of educational applications, including math problem retrieval (Reusch et al., 2021) , problem recommendation (Liu et al., 2018) , and prob... | 0 |
In this paper we employ the theory of frame semantics as a meaning representation framework for dialogues from the domain of disaster response. Our work is part of a larger research project developing methods to capture and interpret verbal team communication in disaster response scenarios and use the extracted run-tim... | 0 |
Understanding of narrative text requires the ability to read between the lines and determine changes in the characters' emotional states concerning the ongoing situation described in the text. Given this motivation, Rashkin et al. (2018) proposed a framework for annotating fully-specified chains of mental states with r... | 0 |
Mushrooming as a scalable lifelong learning paradigm, Massive Open Online Courses (MOOCs) have enjoyed significant limelight in recent years, both in industry and academia (Haggard et al., 2013) . The euphoria is about the transformative potential of MOOCs to revolutionize online education (North et al., 2014) , by con... | 0 |
L'objectif des Systèmes de Dialogue Oral Homme-Machine (SDOHM) est de rendre la machine capable de dialoguer d'une manière naturelle avec l'Homme. Pour ce faire, ces systèmes doivent être dotés d'un gestionnaire de dialogue qui occupe une place très importante dans ce type de systèmes. Le domaine de la gestion de dialo... | 0 |
Content selection is a fundamental task in conceptto-text generation (Reiter and Dale, 2000) . A practical generation system typically operates over a large database with multiple entries that could potentially be included in a text. A content selection component determines what subset of this information to include in... | 0 |
Modern search engines for text documents typically employ multi-stage ranking. The first retrieval stage extracts top candidate documents matching a query from a large search index with a simple ranking method. The second stage or a later stage uses a more complex machine learning algorithm to re-rank top results thoro... | 0 |
Opaque contexts have been an active area of research in natural language semantics since Frege's original discussion of the puzzle (Frege, 1892) . A sentence like (1) has a non-contradictory interpretation despite the fact that the two referring expressions Hesperus and Phosphorus refer to the same entity, the planet w... | 0 |
Properties can be mapped to features in a machine learning algorithm in different ways, potentially yielding different results (see, e.g., Hu and Kibler 1996 and Pagallo and Haussler 1990) . This paper investigates interactions between collocational properties and methods for organizing them into features. Collocations... | 0 |
Natural Language Inference (NLI) is a fundamental task in Natural Language Processing in which the objective is to determine if the hypothesis is true (entailment), false (contradiction) or undetermined (neutral), given a premise. Entailment, Contradiction and Neutral (semantic independence) are semantic concepts that ... | 0 |
Sub-sentential alignments of parallel text play an important role in statistical machine translation (SMT). Aligning parallel data on the word-or phrase-level is typically one of the first steps in building SMT systems, as those alignments constitute the basis for the construction of probabilistic translation dictionar... | 0 |
In all natural languages, components that can be easily deduced by the reader are frequently omitted f~om expressions in texts. In Japanese in particular, the subject and object are often omitted. These phenomena cause problems in machine translation because components not overtly indicated in the source language (i.e.... | 0 |
We examine the task of text to 3D scene generation. The ability to map descriptions of scenes to 3D geometric representations has a wide variety of applications; many creative industries use 3D scenes. Robotics applications need to interpret commands referring to real-world environments, and the ability to visualize sc... | 0 |
Language identification is the task of predicting the language(s) that a given document is written in. It can be seen as a text categorization task in which documents are assigned to pre-existing categories. This research field has found renewed interest in the 1990s due to advances in statistical approaches, and it ha... | 0 |
It is now generally accepted, especially since the work of Lakoff and associates (e.g. [14, 13, 15] ) that much of everyday discourse shows evidence of metaphor. Consequently, the question of how metaphor should be interpreted and what the semantic consequences are of using a metaphor is of major importance in determin... | 0 |
In this paper, we present the joint contribution of Unbabel and IST to the WMT 2021 Shared Task on Metrics. We participated in the segment-level and system-level tracks, as well as the "QE as a Metric" task.Similar to our participation last year (Rei et al., 2020b) , most of the models are based on the COMET framework ... | 0 |
This paper describes an experimental implementa~ tion of a multimodal interface. Specifically, the authors have developed a multimodal drawing tool. The multimodal drawing tool allows users to draw pictures by using multiple modalities; mouse, keyboard and voice input, in various combined ways.Recently, most user inter... | 0 |
Automatic text summarization (Kyoomarsi et al., 2008; Gupta & Lehal, 2010) is reducing the source text into a shorter form retaining its information content and overall meaning. Text Summarization (Lin, 2009) Process can be divided into two phases: 1) Pre Processing phase (Gupta & Lehal, 2011a) is structured representa... | 0 |
Sentiment analysis (Pang and Lee, 2008; Liu, 2012) is a fundamental task in the field of natural language processing. As one of the subtasks of sentiment analysis, the goal of aspect level sentiment classification is to infer the sentiment polarity (e.g. positive, negative, neutral) of the aspect (e.g. food, service). ... | 0 |
For English and a handful of other languages, there are large, well-annotated corpora with a variety of linguistic information ranging from named entity to discourse structure. Unfortunately, for the vast majority of languages very few linguistic resources are available. This situation is likely to persist because of t... | 0 |
Hierarchical phrase-based translation (Chiang, 2007) explores formal synchronous context free grammar (SCFG) rules for translation. Two types of nonterminal symbols are used in translation rules: nonterminal X in ordinary SCFG rules and nonterminal S in glue rules that are specially introduced to concatenate nontermina... | 0 |
In recent years, several algorithms have been developed to acquire semantic lexicons automatically or semi-automatically using corpus-based techniques. For our purposes, the term semantic lexicon will refer to a dictionary of words labeled with semantic classes (e.g., "bird" is an animal and "truck" is a vehicle). Sema... | 0 |
Deep neural networks enabled breakthroughs in natural language generation tasks such as machinetranslation (Zhang and Zong, 2015) , image captioning (Hossain et al., 2019) , and more. Generating the text is done by employing a conditional language model as the decoder component, responsible for predicting the next word... | 0 |
Online communication is an extremely popular form of social interaction. Unlike face-to-face communication, online instant messaging (IM) tools are extremely limited in conveying emotions or the context associated with a communication. Users have adapted to this environment by inventing their own vocabulary, e.g., by p... | 0 |
Historical text normalization is a common approach to making historical documents accessible and searchable. It is a challenging problem, since most historical texts were written without fixed spelling conventions, and spelling is therefore at times idiosyncratic (Piotrowski, 2012) .Traditional approaches to historical... | 0 |
The usage and meaning of words may change over time. For example, the word "gay" originally meant "carefree," but now it also means "homosexual." Such a phenomenon is not unique to English (Hou et al., 2020) . Techniques for quantitatively analyzing the semantic changes of words over time, also known as diachronic chan... | 0 |
The dialogue policy of a dialogue system decides on what dialogue move (also called action) the system should make given the dialogue context (also called dialogue state). Building hand-crafted policies is a hard task, and there is no guarantee that the resulting policies will be optimal. This issue has motivated the d... | 0 |
The semantic concepts of entailment and contradiction are central to all aspects of natural language meaning (Katz, 1972; van Benthem, 2008) , from the lexicon to the content of entire texts. Thus, natural language inference (NLI) -characterizing and using these relations in computational systems (Fyodorov et al., 2000... | 0 |
In this paper, we describe our systems submitted to this year's translation shared tasks at WAT 2017 (Nakazawa et al., 2017) . For this year, we focused on scientific paper (ASPEC Japanese-English, English-Japanese) and newspaper (JIJI Corpus Japanese-English, English-Japanese) translation subtasks.We use a simple Neur... | 0 |
Video has become a ubiquitous way of communication on the Internet, podcast channels, as well as mobile devices. Accelerated by the explosive spread of video data, automatic analysis of semantic video content remains a promising area. The advances of this task can provide comparatively favorable preconditions for subse... | 0 |
In a dialogue system (DS), the dialogue manager (DM) is responsible for choosing the system's contribution to a conversation. Several studies (e.g. (Ultes et al., 2015; Bertrand et al., 2011; Jaksic et al., 2006; Partala and Surakka, 2004)) show that adjusting the system's behaviour to the user can improve the user exp... | 0 |
In order to cope with information available in many languages, modern information systems need large, high quality and multilingual lexical resources. Building such a resource is very expensive. To reduce these costs, we chose to use the "collaborative" development paradigm already used with LINUX and other open source... | 0 |
While in English the "lexical bottleneck" problem (Briscoe, 1991) seems to be softened (e.g. WordNet (Miller, 1990) , Alvey Lexicon (Grover et al., 1993) , COMLEX (Grishman et al., 1994) , etc.) there are no available wide range lexicons for natural language processing (NLP) for other languages. Manual construction of ... | 0 |
Latent Dirichlet Allocation (LDA) (Blei et al., 2003) , like its more established competitors Latent Semantic Indexing (LSI) (Deerwester et al., 1990) and Probabilistic Latent Semantic Indexing (PLSI) (Hofmann, 1999) , is a model which is applicable to the analysis of text corpora. It is claimed to differ from LSI in t... | 0 |
While neural network based machine learning models serve as the backbone of many text and image processing systems, recent studies have shown that they are vulnerable to adversarial examples. Traditionally, an adversarial example is a sample from the classifier's input domain which has been perturbed in such a way that... | 0 |
There is increasing amount of work on computational modeling of Urdu language. As various groups work on the language, diversity in analysis is also developed. In this context, there has been some work on Urdu part of speech (POS) tagging, which has caused multiple tagsets to appear. Thus, there is also need to converg... | 0 |
Word sense ambiguity is a major hurdle for accurate information extraction, summarization and machine translation. The utility of Word Sense Disambiguation (WSD) depends on the accuracy and on how useful the sense distinctions are. The first issue is quantitative, as it can be measured using a WSD system on certain dat... | 0 |
Since it was first introduced to the NLP community by Dunning (1993) , the G 2 log-likelihood-ratio statistic 1 has been widely used in statistical NLP as a measure of strength of association, particularly lexical associations. Nevertheless, its use remains controversial on the grounds that it may be unreliable when ap... | 0 |
Most machine translation systems use sequential decoding strategies where words are predicted one-by-one. In this paper, we present a model and a parallel decoding algorithm which, for a relatively small sacrifice in performance, can be used to generate translations in a constant number of decoding iterations.We introd... | 0 |
Advances in conversational assistants have helped to improve the usability of smart speakers and consumer wearables for different tasks. Semantic parsing is one of the fundamental components of these assistants and it helps to convert the user input in natural language to a structure representation that can be understo... | 0 |
Automated data mining of patents has had a long history of research, driven by the large volume of patents produced each year and the many tasks to which they are put to use, including prior art investigation, competitive analysis, and trend detection and forecasting (Tseng, 2007) . Much of this work has concentrated o... | 0 |
Everything starts with a copy & paste and, of course the flood of documents that we see today could not exist without the practical ease of copy & paste. This is not new but what is new is that the availability of archives allows us to study a vast amount of papers in our domain (i.e. Natural Language Processing, NLP, ... | 0 |
It is a well-known fact that definition queries are very popular across users of commercial search engines (Rose and Levinson, 2004) . The essential characteristic of definition questions is their aim for discovering as much as possible descriptive information about the concept being defined (a.k.a. definiendum, pl. de... | 0 |
Action recognition is the task of identifying the action being depicted in a video or still image. The task is useful for a range of applications such as generating descriptions, image/video retrieval, surveillance, and human-computer interaction. It has been widely studied in computer vision, often on videos (Nagel, 1... | 0 |
Chinese is one of the oldest written languages in the world, but it does not attract much attention in top NLP research forums, probably because of its peculiarities and drastic differences from English. There are sentences, words, characters in Chinese, as illustrated in Figure 1 . The top row is a Chinese sentence, w... | 0 |
In recent times, text has become a preferred mode of communication (Reporter, 2012; ORTUTAY, 2018) over phone/video calling or face-to-face communication. New challenges and opportunities accompany this change. Identifying sentiment from text has become a sought after research topic. Applications include detecting depr... | 0 |
The Arab World is a vast geographical region that covers North Africa and Southwest Asia, boasting a population of around 400M that speak different derivatives of a common language. However, by virtue of its size and cultural variety, there exists a dialect continuum across the region wherein the language varieties of ... | 0 |
With the development of human-machine interaction (HMI) applications, textual dialogue scenes appear more frequently. These scenes request effective and high-performance emotion recognition systems helping in building empathetic machines (Young et al., 2018) . Therefore, emotion recognition in dialogues (ERD) is gettin... | 0 |
Knowledge discovery is a field of computer science that shows an accelerated growth in the past three decades. Advances in this area have been applied in many domains, from databases (Fayyad et al., 1996; Stahl et al.) to images (Lu et al., 2016) and natural language text (Carlson et al., 2010) . Specifically in natura... | 0 |
Semantic parsing, the task of assigning to a natural language expression a machine-interpretable meaning representation, is one of the most difficult problems in NLP. A meaning representation of a document will describe who did what to whom, when, where, why and how in the context. This problem is well studied in NLP, ... | 0 |
Coherence, what makes dialogue utterances unified rather than a random group of sentences, is an essential property to pursue an open-domain Figure 1 : An illustrative example of how our GRADE evaluates dialogue coherence by incorporating graph information on topic transitions from a commonsense graph. Topic keywords o... | 0 |
Arabic Dialects is one of three variations of Arabic language (Classical Arabic CA, Modern Standard Arabic MSA and dialectal Arabic DA). It is a native language of Arabic people used in the communication among them and in the social media (Itani, 2018) . Each country of 21 countries, in the Arab world, has its own dial... | 0 |
Tweets are short messages shared on Twitter, one of the most popular social networking services with 326 million monthly active users word wide (Twitter, 2018) . Tweets often use specialized language such as abbreviations (e.g., TBH: To be honest), hashtags (e.g., #NBAFinals), emoticons and emojis. The Oxford Dictionar... | 0 |
The performance of machine translation and summarization has been approaching a near-human level in virtue of pre-trained encoder-decoder models, such as BART (Lewis et al., 2020) and T5 (Raffel et al., 2020) . The same technology has been applied to dialogue systems, which are now expected to be put to practical use.T... | 0 |
All transport systems have evolved communication procedures with similar characteristics to deal with a common problem; how to ensure that moving vehicles (aircraft, ships, trains) proceed on course, safely and efficiently. These two criteria (safety and efficiency) are, inevitably, in opposition to each other. It woul... | 0 |
How similar are two corpora? The question arises on many occasions. Does it matter whether lexicographers use this corpora or that, or are they similar enough for it to make no difference? (The original impetus for the research was the question, "are the fiction and jourDali~m parts of the Longman Lancaster Corpus and ... | 0 |
Contrairement aux documents nés numériques, les textes historiques sont les résultats d'une numérisation massive des ressources historiques. Étant donné que cette quantité de documents analogiques est encore assez importante malgré la récente évolution vers la création de documents numériques, des efforts considérables... | 0 |
The main paradigm for adapting pretrained models for classification (Radford, 2018; Dong et al., 2019; Devlin et al., 2018) is fine-tuning via an explicit classifier head. However, an alternative approach has arisen: adapting the pretrained language model directly as a predictor through autoregressive text generation (... | 0 |
Knowing when an event happened is useful for a lot of use cases. Examples are in the fields of time-aware information retrieval, text summarization, automated timeline generation, and automatic knowledge base population. Many facts in a knowledge base are only true for a certain time period, for example the presidency ... | 0 |
Recently, to build large structural knowledge bases (KB), great efforts have been made on extracting relational facts from natural language texts. A relational fact is often represented as a triplet which consists of two entities (an entity pair) and a semantic relation between them, such as < Chicago, country, U nited... | 0 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.