Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18265 papers
The present study aims at comparing the effects of two subtypes of cognitive reappraisal (i.e., stimulus-focused vs. goal-based reappraisal) to reduce anticipatory anxiety of pain. Affective ratings, startle reflex, and autonomic measures (electrodermal and heart rate changes) were used as a measure of emotion regulation success. A total of 86 undergraduate students completed an anticipatory task in which they had to regulate their negative emotions or react naturally when faced with the possibility of receiving a painful thermal stimulus. Participants were randomly assigned to two experimental groups to compare the stimulus-focused and goal-based strategies explored here. Our results revealed enhanced self-reported anxiety, electrodermal activity and eyeblink response when participants tried to voluntarily down-regulate their negative emotions, compared to the control instruction. Differences between both cognitive reappraisal groups were not found. These unexpected findings suggest that brief reappraisal instructions may not necessarily be favorable for regulating emotions during anticipation of aversive events. Moreover, these results are further explained in terms of the pain expectation, the painful stimuli modality, and emotion regulation instructions.
The application value of the convolutional neural network (CNN) algorithm in the diagnosis of sports knee osteoarthropathy was investigated in this study. A network model was constructed in this experiment for image analysis of magnetic resonance imaging (MRI) technology. Then, 100 cases of sports knee osteoarthropathy patients and 50 healthy volunteers were selected. Digital radiography (DR) images and MRI images of all the research objects were collected after the inclusion of the two groups. Besides, the important physiological representations were extracted from their image data graphs, and the hidden complex relationships were learned. The state without input results was judged through convolutional network calculation, and the result prediction was given. On this basis, there was an analysis of the diagnostic efficiency of traditional DR images and MRI images based on CNN for patients with sports knee osteoarthropathy. The results showed that the MRI images analyzed by the CNN model showed a more obvious display rate than DR images for some nonbone changes of osteoarthritis. The correlation coefficient between MRI image rating and visual analog scale (VAS) was 0.865, which was higher than 0.713 of DR image rating, with a statistical meaning ( <math xmlns="http://www.w3.org/1998/Math/MathML" id="M1"> <mi>P</mi> <mo><</mo> <mn>0.01</mn> </math> ). For cases with mild lesions, the number of cases detected by MRI based on CNN algorithm in 0–4 image rating was 15, 18, 10, 6, and 7, respectively, which was markedly better than that of DR images. In short, the MRI examination based on the CNN image analysis model could extract important physiological representations from the image data and learn the hidden complex relationships. The convolutional network was calculated to determine the state of the uninput results and give the result predictions. Moreover, MRI examination based on the CNN image analysis model had high overall diagnostic efficiency and grading diagnostic efficiency for patients with motor knee osteoarthropathy, which was of great significance in clinical practice.
The aim of this article is to identify the Old English exponent for the semantic prime LIVE following the principles of the Natural Semantic Metalanguage theory (Wierzbicka 1996, Goddard & Wierzbicka 2002, Goddard 2011). The methodology applied in the study is based on previous research in Old English semantic primes. In these terms, a search for those Old English words conveying the meaning of the semantic prime LIVE is made. This search selects the verbs (ge)buan, drohtian, (ge)eardian, (ge)libban, and wunian as candidate words for prime exponent. Then, these verbs are analysed in terms of morphological, textual, semantic, and syntactic criteria. With this purpose, relevant information on these words has been gathered from different lexicographical and textual sources in Old English, such as the Dictionary of Old English, the Dictionary of Old English Corpus, and the lexical database of Old English Nerthus. After the analysis of these verbs, the conclusion is drawn that the Old English verb (ge)libban is selected as prime exponent, as it satisfies the requirements proposed by each criterion.
This paper presents a full procedure for the development of a Part-of-Speech (POS) tagged corpus of Old Catalan. As an extremely low-resource language with rich inflection and frequent homographs, Old Catalan poses non-trivial problems in the development of a searchable constituency-based treebank. We demonstrate, however, that a semi-supervised method of incrementally building training data using both neural and memory-based taggers, together with the Pyrrha annotation tool is highly efficient and yields accurate results. We propose that this simple and effective method could easily be extended to other low-resource historical languages for which no NLP tools exist yet.
This paper compares two influential theories of processing difficulty: Gibson (2000)'s Dependency Locality Theory (DLT) and Hale (2001)'s Surprisal Theory. While prior work has aimed to compare DLT and Surprisal Theory (see I compare estimated surprisal values from two models, an RNN and a Transformer neural network, as well as DLT integration cost from a hand-parsed treebank, to reading times from the Dundee Corpus. The results for integration cost corroborate those of Ultimately, I conclude that a broad-coverage model must integrate both theories in order to most accurately predict processing difficulty.
Recently, Chinese Syntax-aware Semantic Role Labeling (SRL) has attracted the attention of many researchers because syntax information is related to semantic role labeling intuitively. In this paper, we integrate constituency representation into SRL by combining multiple methods into a unified model. Specifically, our proposed model combine three methods for integrating constituent into SRL, including pipeline (use syntax structure), hard parameter (share encoder), implicit representation integration (not share encoder). We verify the effect of our model on Chinese Proposition Treebank (CPB) 1.0 dataset and conduct some ablation experiments to verify the impact of various parts of the model.
OBJECTIVE: To evaluate the effect magnitude of different parameters on smile attractiveness. MATERIALS AND METHODS: A reference and 13 images were produced by manipulating 13 parameters. Image rating was performed with a 4-point Likert scale from least attractive (1) to most attractive (4). Image raters included laypeople, dental students, dentists, and dental specialists. Friedman and Wilcoxon image were used for estimate of effect size. Parameters were classified into small (0.10- < 0.30), medium (0.30- < 0.5), or large (≥0.50). RESULTS: A total of 1040 people participated with good consistency (α = 0.861), and moderate reliability (0.64-0.7). The reference image had the highest rank (laypeople:11.79, dental background: 12.55). For effect size; gingival margin level (-0.11, -0.01), teeth width proportion (-0.09, -0.10), inverted smile arch (-0.09, -0.21), commissure line cant (-0.15, -0.17) and low smile (-0.24, -0.23) had small effect size; occlusal plane cant (-0.36, -0.49), midline cant (-0.36, -0.48), and midline shift (-0.37, -0.49) had medium effect size; diastema (-0.55, -0.54) and color (-0.56, -0.56) had large effect size for the laypeople and dental groups. High smile (-0.42, -0.51), incisor edge symmetry (-0.46, -0.54) had medium effect size in laypeople group and large effect size in the dental group. Width to length tooth proportion (-0.26, -0.39) had small effect size in the laypeople group and medium effect size in the dental group. CONCLUSIONS: Smile parameters had different effect magnitude on smile attractiveness and were classified into small, medium, or large parameters. Neither laypeople nor professionals have a collective judgment on what constitutes a beautiful smile. CLINICAL SIGNIFICANCE: This study investigated the effect magnitudes of 13 smile parameters and presented a small, medium, and large smile parameters classification. It should provide the clinician with an insight into the expected effect each parameter has on the smile.
The article focuses on the linguistic personality of A.V. Suvorov – one of the most important historical figures famous for his military achievements, including his service and warfare in Slavic territories (military service in the Lublin region, participation in hostilities in Poland, the Battle of Brest, etc.), a prominent representative of the Russian cultural elite in the 18th century. The study employs methods of linguopersonology and historical lexicology. It aims to describe a historical language personality, viewed as a reflection of both an individual with a linguistic potential, a type of specific linguistic reflection, and of the era as a whole with its inherent linguistic features and tendencies. Processes in the language at a given time period are also manifested in the language of an individual. The study focuses on the borrowed lexis, reflecting the general instability of the Russian language system of the period, which resulted from the prevailing multilingualism of the Russian nation, a significant increase in foreign words in the Russian language, and a change in linguistic norms. Through the prism of xenolexis, the authors describe Suvorov’s language personality reflected in his letters to note his broad outlook, fluency in native and foreign languages, linguistic intuition, innovative use of units of native and foreign languages.
Search is one of the key functionalities in digital platforms and applications such as an electronic dictionary, a search engine, and an e-commerce platform. While the search function in some languages is trivial, Khmer word search is challenging given its complex writing system. Multiple orders of characters and different spelling realizations of words impose a constraint on Khmer word search functionality. Additionally, spelling mistakes are common since robust spellcheckers are not commonly available across the input device platforms. These challenges hinder the use of Khmer language in search-embedded applications. Moreover, due to the absence of WordNet-like lexical databases for Khmer language, it is impossible to establish semantic relation between words, enabling semantic search. In this paper, we propose a set of robust solutions to the above challenges associated with Khmer word search. The proposed solutions include character order normalization, grapheme and phoneme-based spellcheckers, and Khmer word semantic model. The semantic model is based on the word embedding model that is trained on a 30-million-word corpus and is used to capture the semantic similarities between words.
This paper examines the use of Italian digital language, which is often evaluated in negative terms. Considering the fact that internet communication occupies an important place in the life of modern man, the study of the features of digital language has been the subject of much research. For those born in the digital age (it. nativi digitali), digital has become the norm to the extent that it is difficult to imagine life without multimedia interaction through modern means of communication (Bralić 145). Digital text is different from traditional written text and the rapid obsolescence of new media is changing the habits of digital language users. Italian, which has existed exclusively in the traditional written form for centuries, and has received full spoken use in the last seventy years (largely thanks to television), faces today a new revolutionary phase of development in which the majority of Italians in everyday life use written digital language. In this way, the digital age marked a return to the Italian written language. However, the language of forums and social networks is an informal language (e-Italian), quite different from the former, exceptionally formal, written Italian. The aim of this paper is to study and explain the linguistic features of the Italian language in Internet communication. The focus is on the language of blogs, forums, and social networks written in Italian over the last three years, from the beginning of 2018 to the end of 2020. The question is whether everything that deviates from the norm in the language is wrong or if, on the contrary, demonstrates the stability and ability of the language to adapt to new media and thus new conditions. The major changes on social networks are the result of the transition from the elite use of the network to the “mass network” (Gheno 2017, 103). The changes are also heading towards the direction that has yet to be identified. Thus, we notice that the use of certain language features on social networks such as abbreviations, acronyms and other similar phenomena was a way of distinction, but also a necessity dictated by technical limitations such as restricted space for writing messages and the high cost of network connection. Therefore, it comes to no surprise that in recent years we have witnessed a writing normalization directed towards approximating some kind of linguistic norm. Finally, after having removed the space and time limitations and as a result of the possibility of spell checking that is suggested by smart devices while writing, even the so-called “language play and use of creative forms of writing” has become practically a waste of time. The fact that we are in the normalization phase can also be seen thanks to other novelties on social networks. One of them is caused by the policy of some platforms that is aimed at using one’s own name and abandoning the nickname, leading to an interesting social effect demonstrating that haters do not necessarily hide behind nicknames. Moreover, there is a tendency to give more importance to the interlocutor who signs with his own name, as contrasted with those who use a nickname. It becomes normal again to introduce yourself by your real name and surname, without leaving the impression of a person that is hidden behind a mask or nickname. The use of language on social networks has changed thoroughly over time and continues to change even today, both in Italian and in other languages. It is highly probable that over time users will pay more attention to the impression they leave online and, thus, be more careful when it comes to the language, they use by respecting the prescribed language norms. In addition to dealing with language dilemmas, it is necessary to establish the right habits that will allow us to live a comfortable life online and accept the fact that we have become like mini public figures who are responsible for what they say. We should also keep in mind that, on social networks, the most emphasized part of our online personality is presented primarily by words.
<p><strong>Purpose: </strong></p> <p>The current study explores the spillover effects of offensive commenting in online community from the lens of emotional and behavioral contagion. Specifically, it examines the contagion of swearing –a linguistic mannerism that conveys high arousal emotion –based upon two mechanisms of contagion: mimicry and social interaction effect.</p> <p><strong>Design/methodology/approach:</strong></p> <p>The study performs a series of mixed-effect logistic regressions to investigate the contagious potential of offensive comments collected from YouTube in response to Donald Trump’s 2016 presidential campaign videos posted between January and April 2016.</p> <p><strong>Findings:</strong> </p> <p>The study examines non-random incidences of two types of swearing online: public and interpersonal. Findings suggest that a first-level (a.k.a. parent) comment’s public swearing tends to trigger chains of interpersonal swearing in the second-level (a.k.a. child) comments. Meanwhile, among the child-comments, a sequentially preceding comment’s swearing is contagious to the following comment only across the same swearing type. Based on the findings, the study concludes that offensive comments are contagious and have impact on shaping the community-wide linguistic norms of online user interactions.</p> <p><strong>Originality/value:</strong> </p> <p>The study discusses the ways in which an individual’s display of offensiveness may influence and shape discursive cultures on the Internet. This study delves into the mechanisms of text-based contagion by differentiating between mimicry effect and social interaction effect. While online emotional contagion research to this date has focused on the difference between positive and negative valence, Internet research that specifically look at the contagious potential of offensive expressions remain sparse.</p>
Modern mass media, as an important means of informing the public, determines people’s consciousness, shapes their interests and determines the mood. It has also successfully incorporated the role of an educator from the very beginning. Thus, the media linguistics plays a special role in spreading the native Georgian language, as well as in raising the literacy rate of the population. Language is the main tool to guide human cognitive activity. Although, there are many examples of media trying to comply with linguistic norms, modern broadcasting fails to maintain proper Georgian; digital media being particularly full of many linguistic and stylistic errors. It should also be noted that electronic media is the only source of information for many, especially the young people. Given the high social importance, journalists should be careful with dealing with the language. The development and transformation process of mass media in the global information space is not only indirectly but also directly reflected, especially in the formation of the Internet media linguistics. The current social and political changes in the country have also affected the process of language deformation. The scientific article deals with the influence of media language on the linguistic features of the Georgian society. It discusses the problematic issues of modern publicist thinking, focusing on the linguistic distortions that modern journalists are characterized with. The speech problem of media representatives is topical, complex and large-scale. Their speeches are the basis of the communicative interaction of the society, contribute to the linguistic influence, national identity, mutual understanding of people, perception of the world. The media significantly influences the value system, mentality and literary norms used in any society. Violation of literary norms, which is observed in modern Georgian media, has a negative impact on the audience and the level of literary competence of the speakers.
A disponibilidade de recursos para o processamento computacional constitui um dos fatores de sobrevivência de uma língua. O objetivo deste trabalho foi implementar um fragmento do nheengatu no formalismo Grammatical Framework, especialmente projetado para o desenvolvimento de aplicações multilíngues. Outrora mais falado que o português na Amazônia, o nheengatu está ameaçado de extinção, embora ainda conte com estimados 14000 falantes. O fragmento restringe-se a orações que expressam estados contingentes e não-contingentes, mas inclui fenômenos gramaticais estruturalmente complexos típicos da família tupi-guarani, os quais contrastam fortemente com as construções equivalentes em português e inglês. Constitui um dos módulos da GrammYEP, uma gramática computacional multilíngue que integra módulos análogos do inglês e do português. A implementação tomou como ponto de partida as descrições gramaticais não formalizadas de Navarro (2011) e Cruz (2011). A formalização revelou lacunas e inconsistências nessas abordagens, em parte sanados por meio de uma reanálise dos dados. A GrammYEP alcançou resultados bastantes satisfatórios na tradução do e para o nheengatu. Traduziu para o português e o inglês a totalidade de um conjunto-teste de 142 sentenças dessa língua. Inversamente, verteu para o nheengatu 98,18% e 84,11% dos conjuntos-teste correspondentes em português e inglês. Por outro lado, analisou apenas dois exemplos de um conjunto-teste negativo com 171 construções agramaticais em nheengatu. Desta avaliação resultou um treebank com 243 sentenças do nheengatu, emparelhadas com as sentenças equivalentes em português e inglês.
Abstract Classical views suggest that experienced affect is related to a specific bodily response ( Fingerprint hypothesis ), whereas recent perspectives challenge this view postulating that similar affective experiences rather evoke different physiological responses. To further advance this debate in the field, we used representational similarity analysis (N= 64) to investigate the correspondence between subjective affect (arousal and valence ratings) and physiological reactions (skin conductance response [SCR], startle blink response, heart rate and corrugator activity) across various emotion induction contexts (picture viewing task, sound listening task and imagery task). Significant similarities were exclusively observed between SCR and arousal in the picture viewing task. However, none of the other physiological measures showed a significant relation with valence and arousal ratings in any of the tasks. These findings tend to support the populations hypothesis, suggesting that there is no clear match between the evoked physiological responses and the experienced subjective affect between individuals. Statement of relevance The subjective affective experience evoked by an event is accompanied by physiological responses. The correspondence between physiological response patterns and the experienced affect, however, is still under debate. Classical views ( Fingerprint hypothesis ) suggest that affect is related to a specific physiological response, whereas recent perspectives ( Populations hypothesis ) challenge this view, postulating rather different physiological responses. In the current study, we used representational similarity analysis (RSA) to examine the relation between affective experience, assessed using valence and arousal ratings, and the evoked physiological reactivity across three affect-inducing contexts. Results showed significant similarities exclusively between SCR and arousal in the passive picture viewing task. However, none of the other physiological measures showed a significant relation with valence and arousal ratings in any of the tasks, supporting the populations hypothesis. These findings invite to reframe the relation between physiology and affect from invariant and homogeneous to variant and context-dependent.
To the Editor: Coronavirus disease 2019 (COVID-19), which broke out in 2019, has become a global pandemic. Similar to severe acute respiratory syndrome coronavirus (SARS-CoV) in 2003, SARS-CoV-2 could cause acute lung injury and cytokine storms characterized by the increased interleukin (IL)-8, IL-6, and tumor necrosis factor α (TNF-α).[1] Perspective studies in those survivors from the severe acute respiratory syndrome (SARS) epidemic in 2003 revealed that those SARS patients manifested varying degrees of pulmonary interstitial fibrosis.[2] Similarly, patients with severe COVID-19 are also featured by the diffuse alveolar damage along with alveolar interstitial fibrosis.[3] Pirfenidone (Beijing Contini Pharmaceutical Co., Ltd, Beijing, China) can inhibit the biological activity of fibroblasts and reduce matrix collagen deposition, prevent inflammasome activation and limit oxidative stress responses, supporting a therapeutic potential against idiopathic pulmonary fibrosis (IPF).[4] Given the presence of alveolar interstitial fibrosis in severe COVID-19 patients and the effect of pirfenidone on anti-inflammatory responses and anti-fibrosis, we hypothesized that pirfenidone can play a positive role in COVID-19 patients, thereby reducing the incidence of complications following SARS-CoV-2 infection. We thus conducted a clinical trial to assess the potential therapeutic effect of pirfenidone on severe COVID-19 patients. This trial (ClinicalTrials.gov number: NCT04282902; Chinese Clinical Trial Register number: ChiCTR2000030333) was conducted from January 31 to March 3, 2020 at Tongji Hospital (Headquarters Campus, Caidian Campus, Guanggu Campus) and Jingzhou Hospital (Hubei, China), which was approved by the Institutional Review Board of Tongji Hospital and Jingzhou Hospital. Male and non-pregnant female COVID-19 patients (≥18 years) with a blood oxygen saturation (SaO2) of 94% or less, and a ratio (PaO2:FiO2) of ambient air or partial oxygen pressure (PaO2) to inhaled oxygen (FiO2) of 300 mmHg or less, were eligible for the study. The exclusion criteria were: patient disinterest in the study; the presence of conditions that did not allow for safe compliance, including hypersensitivity to pirfenidone; liver disease (eg, alanine aminotransferase levels >5 times the upper limit of the normal range [ULN] or aspartate aminotransferase levels >5 times ULN; contraindications of pirfenidone and pre-existing interstitial lung disease [ILD]). Consecutive patients, who meet the inclusion criteria, were randomly assigned in a 1:1 ratio to pirfenidone (200 mg, three times daily for the first two days and 400 mg, three times daily thereafter) plus standard therapy or standard therapy alone. Pirfenidone was given through a nasogastric tube in patients who were unable to swallow. The primary end-point was the absolute changes from baseline in the total score on the King's Brief Interstitial Lung Disease (K-BILD) questionnaire at the 4th week, a change between 4 and 8 points has been suggested to represent a meaningfulchange. Secondary endpoint was the absolute change in computed tomography (CT) values, the total CT value was the sum of individual lobe values and ranged from 0 (no involvement) to 25 (maximum involvement). Other secondary outcomes included clinical laboratory findings (cytokines, biochemical indicators, etc) and the proportion of patients with clinical improvement. Safety outcomes included adverse events that occurred during treatment and premature discontinuation of treatment. Adverse events were classified according to the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0 [Supplementary methods in Supplementary materials, https://links.lww.com/CM9/A929]. A total of 146 COVID-19 patients were recruited, 124 of which were from Tongji Hospital and the rest 22 were from the Central Hospital in Jingzhou. Seventy-three patients were randomly assigned for pirfenidone treatment, and the remaining 73 patients received the standard treatment alone [Supplementary Figure 1, https://links.lww.com/CM9/A647]. The median age of patients was 62.0 years (interquartile range [IQR] 53.5–68.5 years), and 64.38% of patients were males. The median interval time between symptom onset and randomization was 40 days (IQR, 25–50 days). At the time of admission, there were no significant differences between the two groups in terms of demographic characteristics, basic laboratory assays, clinical treatment K-BILD scores, and CT scores [Supplementary Table 1, https://links.lww.com/CM9/A646]. Although the difference of K-BILD scores did not reach a statistical significance after treatment (75.93 ± 10.07 vs. 76.33 ± 9.15, P = 0.911), a trend for the increase from baseline in patients following a 4-week of pirfenidone treatment was noted as compared to that of patients assigned in the standard treatment group (ΔK-BILD, 26.53 ± 11.12 vs. 22.73 ± 8.00; 3.80 [95% confidence interval, CI = −4.87 to 12.47]) [Supplementary Table 2, https://links.lww.com/CM9/A646]. Similarly, there was no significant difference between two groups in terms of CT images (P = 0.745) after a 4-week of treatment, but some score changes including consolidation (0.30 ± 0.65 vs. 1.07 ± 1.17, P = 0.007), GGO (−12.27 ± 5.72 vs. −11.57 ± 4.07; between-group difference = −0.70, 95% CI = −2.97 to 1.57), and reticulation (−0.90 ± 5.26 vs. −0.30 ± 6.98; between-group difference = −0.60, 95% CI = −3.40 to 2.20) were observed, which reflected the improvement of lung inflammation and interstitial changes [Supplementary Table 3, https://links.lww.com/CM9/A646 and Supplementary Figure 2, https://links.lww.com/CM9/A648]. The levels of pulmonary inflammatory cytokines or coagulopathy biomarker from baseline to the 4th week after receiving treatment were significantly decreased in the pirfenidone group as compared to those from the standard care group, such as IL-2R (−299.00, 95% CI = −430.50 to −105.00, P = 0.010), TNF-α (−3.50, 95% CI = −5.00 to −0.10, P = 0.049), and D-Dimer (−4.57, 95% CI = −8.98 to −0.16, P = 0.021) [Supplementary Table 2, https://links.lww.com/CM9/A646]. In addition, the duration of patients in pirfenidone group from randomization to hospital discharge and in intensive care unit was reduced by 2 days (11.21 ± 10.06 vs. 13.21 ± 16.63 days, −2 days, 95% CI = −9.27 to 5.27; 19.00 [IQR 15.00–22.00] vs. 22.00 [IQR 16.50–25.50]; −2 days, 95% CI = −3.50 to 8.50) [Supplementary Table 2, https://links.lww.com/CM9/A646]. No significant difference was noted for other outcomes such as clinical improvement time, duration of oxygen therapy, and time from randomization to death. However, all patients survived in the pirfenidone treatment group, and two patients were declined to death in the standard care group. The proportional distribution of primary endpoint categories at days 1, 7, 14, and 28 in each patient was presented in Supplementary Figure 3, https://links.lww.com/CM9/A649 and Supplementary Table 3, https://links.lww.com/CM9/A646. The percentages of patients with any adverse event or serious adverse event were similar between the patients from both groups. Among patients with adverse events, 11% (8/73) of patients reduced the dose of pirfenidone and 3% (2/73) of patients discontinued. Among those eight patients with pirfenidone reduced-dose, four cases were reduced to 600 mg/day due to gastrointestinal discomfort and the remaining four cases were reduced to 600 mg/day due to rash. The most common adverse event was diarrhea, which was reported in 11 out of 73 (15%) patients from the pirfenidone group. Some patients have elevated alanine aminotransferase and alanine aminotransferase level [Supplementary Figure 3 and Supplementary Table 2, https://links.lww.com/CM9/A646, https://links.lww.com/CM9/A649]. Given that the COVID-19 is a type of infectious disease that could be transmitted through the respiratory tract, lung function test was not included in this study. According to previous studies, patients infected with SARS-CoV-2 are far more likely to form an interstitial change in the lung, while our statistical indicators including K-BILD and CT image ratings reflected improved situation in terms of interstitial changes in pirfenidone treated patients. However, we failed to observe a significant difference between the two groups both for K-BILD and CT scores. Since our observation period only lasted for 4 weeks, our trial did not yield a significantly positive result, and the observation time could be a major factor. It was noted that pirfenidone did not improve fibrosis, but it did not make the disease worse. Nevertheless, pirfenidone did manifest a strong effect on mitigating the cytokine storm, which seems to be responsible for the complications in severe COVID-19 patients. Indeed, comparative analysis revealed that the levels of IL-2R and TNF-α were decreased significantly following pirfenidone administration. Although the anti-inflammatory effect has not been widely appreciated, pirfenidone has been shown to downregulate inflammatory pathways and the compound may have considerable potential to be deployed as a non-steroidal anti-inflammatory agent.[5,6] To our surprise, our study also found that pirfenidone could significantly decrease the level of D-Dimer, which is relevant to the coagulopathy in the blood. There is evidence that COVID-19 renders patients with an increased risk for acute pulmonary embolism, and anticoagulant therapy might be associated with improved outcomes in patients with severe COVID-19.[7] Therefore, treatment of COVID-19 with pirfenidone may have the potential to reduce the incidence of thrombosis complications. The safety and side-effect profile of pirfenidone in patients with severe COVID-19 associated ILD was similar to that observed in patients with IPF.[8] Despite the presence of certain side effects, but they could be easily managed with supporting therapies and temporary dose reductions or discontinuation. No fatal events were reported in our study, confirming the good safety profile of this drug even in fragile patients. Our study also has several limitations. First, our sample size is limited. Second, although we have tried to avoid bias as much as possible during the trial, while this possibility cannot be completely excluded. Third, our trial lacks of dynamic clinical and laboratory data such as immune cell subsets and so on. Finally, our trail only lasted a 4-week of observation time, which could be a factor to confirm the antifibrotic effect of pirfenidone. In addition, glucocorticoids could also be a potential factor influencing the results of the study, even though there was no significant difference in the initial dosage, total dosage, and duration of treatment between the two groups. Although pirfenidone has not been found to significantly improve the interstitial changes in severe COVID-19 patients, the trial, however, confirmed the benefits of pirfenidone therapy in anti-inflammatory responses, and obtained feasible evidence supporting a potential benefit in anti-thrombotic complications. Collectively, our study found that pirfenidone can be considered as a viable drug to treat patients with severe COVID-19, and confirmed that pirfenidone possesses a good tolerability profile without safety alert. Funding This study was supported by grants from the SARS-CoV-2 Pneumonia Emergency Technology Public Relation Project of Tongji Medical College, Huazhong University of Science and Technology (No. 2020kfyXGYJ043), the National Natural Science Foundation of China (No. 81974456); and the SARS-CoV-2 Pneumonia Emergency Technology Public Relation Project (No. 2020FCA009). Conflicts of interest None.
Current practices to apply temperature scaling assume either a fixed, or a manually-crafted dynamically changing schedule. However, our studies indicate that the individual optimal trajectory for each class can change with the context. To this end, we propose context-aware temperature, a generalized approach to provide an individual optimal temperature trajectory over the context for each vocabulary, while allowing the temperature to be learned along with the remaining model parameters during training. Experiment results confirm that the proposed method significantly improves state-of-the-art language models, achieving a perplexity of 19.90 on Penn Treebank, 33.88 on WikiText-2, and 4.7 on WikiText-103.
As an engineering field, research on natural language processing (NLP) is much more constrained by currently available resources and technologies, compared with theoretical work on computational linguistics (CL). In today’s technology-driven society, it is almost impossible to imagine the degree to which computational resources, the capacity of secondary and main storage, and software technologies were restricted when I embarked upon my research career 50 years ago. While these restrictions inevitably shaped my early research into NLP, my subsequent work evolved, according to the significant progress made in associated technologies and related academic fields, particularly CL.Figure 1 shows the research topics in which I have been engaged. My initial NLP research was concerned with a question answering system, which I worked on during my M.Eng and D.Eng degrees. The research focused on reasoning and language understanding, which I soon found was too ambitious and ill-defined. After receiving my D.Eng., I changed my direction of research, and began to be engaged in processing forms of language expressions, with less commitment to language understanding, machine translation (MT), and parsing. However, I returned to research into reasoning and language understanding in the later stage of my career, with clearer definitions of tasks and relevant knowledge, and equipped with access to more advanced supporting technologies.In this article, I begin by briefly describing my views on mutual relationships among disciplines related to CL and NLP, and then move on to discussing my own research.Language is a complex topic to study, infinitely harder than I first imagined when I began to work in the field of NLP.There is a whole discipline on the study of language—namely, linguistics. Linguistics is concerned not only with language per se, but must also deal with how humans model the world.1 The study of semantics, for example, must relate language expressions to their meanings, which reside in the mental models possessed by humans.Apart from linguistics, there are two fields of science that are concerned with language, that is, brain science and psychology. These are concerned with how humans process language. Then, there are two disciplines in which we are involved—namely, CL and NLP.Figure 2 is a schematic view of these research disciplines. Both of the lower disciplines are concerned with processing language, that is, how language is processed in our minds or our brains, and how computer systems should be designed to process language efficiently and effectively.The top discipline, linguistics, on the other hand, is concerned with rules that are followed by languages. That is to say, linguists study language as a system. This schematic view is certainly oversimplified, and there are subject fields in which these disciplines overlap. Psycholinguistics, for example, is a subfield of linguistics which is concerned with how the human mind processes language. A broader definition of CL may include NLP as its subfield.In this article, for the sake of discussion, I adopt narrower definitions of linguistics and CL. In this narrower definition, linguistics is concerned with the rules followed by languages as a system, whereas CL, as a subfield of linguistics, is concerned with the formal or computational description of rules that languages follow.2CL, which focuses on formal/computational description of languages as a system, is expected to bridge broader fields of linguistics with the lower disciplines, which are concerned with processing of language.Given my involvement in NLP, I would like to address the question of whether the narrowly defined CL is relevant to NLP. The simple answer is yes. However, the answer is not so straightforward, and requires us to examine the degree to which the representations used to describe language as a system are relevant to the representations used for processing language.Although my colleagues and I have been engaged in diverse research areas, I pick up only on a subset of these, to illustrate how I view the relationships between NLP and CL. Due to the nature of the article, I ignore technical details and focus instead on the motivation of the research and the lessons which I have learned through research.Background and Motivation. Following the ALPAC report Pierce et al. (1966), research into MT had been largely abandoned by academia, with the exception of a small number of institutes (notably, GETA at Grenoble, France, and Kyoto University, Japan). There were only a handful of commercial MT systems, being used for limited purposes. These commercial systems were legacy systems that had been developed over years and had become complicated collections of ad hoc programs. They had become too convoluted to allow for changes and improvements. To re-initiate MT research in academia, we had to have more systematic and disciplined design methodologies.On the other hand, theoretical linguistics, initiated by Noam Chomsky (Chomsky 1957, 1965) had attracted linguists with a mathematical orientation, who were interested in formal frameworks of describing rules followed by language. Those linguists with interests in formal ways of describing rules were the first generation of computational linguists.Although computational linguists did not necessarily follow the Chomskyan way of thinking, they shared the general view of treating language as a system of rules. They had developed formal ways of describing rules of language and showed that these rules consisted of different layers, such as morphology, syntax, and semantics, and that each layer required different formal frameworks with different computational powers. Their work had also motivated work on how one could process language by computerizing its rules of language. This work constituted the beginning of NLP research, and resulted in the development of parsing algorithms for context-free language, finite-state machines, and so forth.3 It was natural to use this work as the basis for designing the second generation of MT systems, which was initiated by an MT project (MU project, 1082-1986) led by Prof. M. Nagao (Nagao, Tsujii, and Nakamura 1985).Research Contributions. When I began research into MT in the late 1970s, there was a common view largely shared by the community, which had been advocated by the group of GETA, in France. The view was called the transfer approach of MT (Boitet 1987).The transfer approach viewed translation as a process consisting of three phases: analysis, transfer, and generation. According to linguists, a language is a system of rules. The analysis and generation phases were monolingual phases that were concerned with a set of rules for a single language, the analysis phase using the rules of the source language and the generation phase using the rules of the target language. Only the transfer phase was a bilingual phase.Another view shared by the community was an abstraction hierarchy of representation, called the triangle of translation. For example, Figure 3(a)4 shows the hierarchy of representation used in the Eurotra project, with their definition of each level (Figure 3(b)).By climbing up such a hierarchy, the differences among languages would become increasingly small, so that the mapping (i.e., the transfer phase) from one language to another would become as simple as possible. Independently of the target language, the goal of the analysis phase was to climb up the hierarchy, while the aim of the generation phase was to climb down the hierarchy to generate surface expressions in the target language. Both phases are concerned only with rules of single languages.In the extreme view, the top of the hierarchy was taken as the language-independent representation of meaning. Proponents of the interlingual approach claimed that, if the analysis phase reached this level, then no transfer phase would be required. Rather, translation would consist only of the two monolingual phases (i.e., the analysis and generation phases).However, in Tsujii (1986), I claimed, and still maintain, that this was a mistaken view about the nature of translation. In particular, this view assumed that a translation pair (consisting of the source and target sentences) encodes the same “information”. This assumption does not hold, in particular, for a language pair such as Japanese and English, that belong to very different language families. Although a good translation should preserve the information conveyed by the source sentence as much as possible in the target sentence, translation may lose some information or add extra information.5Furthermore, the goal of translation may not be to preserve information but to convey the same pragmatic effects to readers of the translation.More seriously, the abstract level of representation such as Interface Structure6 in Eurotra focused only on the propositional content encoded in language, and tended to abstract away other aspects of information, such as the speaker’s empathy, distinction of old/new information, emphasis, and so on.To climb up the hierarchy led to loss of information in lower levels of representation. In Tsujii (1986), instead of mapping at the abstract level, I proposed “transfer based on a bundle of features of all the levels”, in which the transfer would refer to all levels of representation in the source language to produce a corresponding representation in the target language (Figure 4). Because different levels of representation require different geometrical structures (i.e., different the of this had to for development of a mathematical of representation with which levels (i.e., to be with their mutual relationships the we to the transfer phase was transfer and Tsujii which was by the of in CL. According to the views of linguists at the a language is an set of expressions in is defined by a set of rules. this number of one generate infinitely of the language. claimed that the of a was by the of its using the rules that the translation the same to translation. That is, the translation of a was by the of its In this of infinitely of the source language could be the translation the translation of a sentence would be by a of a source The translation of a would then be by the of its That is, translation would be in a up from of translation to the mapping of a from the source to the target would be by the of the the for the how to a to the In the project, we called this transfer and Tsujii (Figure with the MT systems, which source expressions with target in an and ad hoc the of transfer in the project was defined and and development of the MT systems from research into CL, more defined and design than MT The project and MT systems the of these design we could not have these in such a of the differences between the of the two disciplines also CL to focus on aspects of language as morphology, syntax, semantics, MT systems must be to all aspects of information conveyed by language. As climbing up a hierarchy that focuses on propositional content does not in good more between CL and NLP is the of of is the single significant in NLP it requires the in which expressions to be to be In other it requires understanding of of are in Figure The Japanese a of in and would be into have and so on the for with transfer, it requires to be (i.e., the in which a to be The nature of made the process of transfer was also a in the analysis which I in the general of CL or linguistics is that it to view language as an system and the of understanding, which requires to or However, NLP require an understanding or of language expressions in of and which may other such as and so I this in the on the of research.Background and Motivation. the I was engaged in MT research, in CL, its early in theoretical linguistics by Chomsky assumed that of of rules the two levels of that is, and surface A way of was also shared by the MT They assumed that climbing up the hierarchy would of which from the representation at one level to another representation at the Because each level of the hierarchy required its own geometrical it was not possible to have a representation, in which representations of all the levels view was changed by the of that used to allow of from one level to it mutual relationships among different levels of representation in a This view was in with our of transfer, which used a bundle of features of different levels for some at the the of That is, structures of all the levels are constrained by the of a and these are encoded in This was also in with our significant development in CL at the same a number of the and the had linguistics and to have significant on research into CL and NLP et al. the NLP of view, the of led to the development of (i.e., for research that would these two to the analysis is, parsing based on Contributions. It is claimed that of In the analysis phase of the up the lower levels of processing could not refer to in levels of representation. This was the main of the of at the early of climbing up the analysis could not refer to that in analysis would the other hand, the could describe at all levels in a single it was possible to refer to at all to down the set of possible in the was still This was we not have ways of and pragmatic linguists were interested in formal ways for and levels of representation, but not so much in how are to be To or pragmatic one may have to refer to the mental models of the (i.e., how humans the or structures single and so These of the of CL research at the main focus is on it is whether or be used as They may be more concerned with the of an than the which an should example, the in for parsing using of and how to the of for models were one of the for and the of an However, models for such as and context-free had to be changed for more complex for such as had to be for were in of describing in a the which was a for treating was very To NLP systems, we had to technologies and processing for at the of to study how we could a into and representations for parsing. The research of an abstract machine for processing of structures and development of a et al. et al. into a more representation, such as and Tsujii et al. and from of structures and models and Tsujii of parsing based on of and their and Tsujii et al. representation of our parsing model is in Figure a sentence, its representation of all the levels was at the stage by using the in the first two The first phase was a that would to in a were from the and a set of were to a in its A would the of for the of The was a which could be in a very and Tsujii This that the surface (i.e., of was used for of second phase was A which also was from the was used to whether of by the first phase could a The did not to whether a could or The second phase of would that could not phase not only the representation of all the but it also extra in the Because the first two phases only use in the the phase would by the first two phases if they to these extra In this the system would to the phases to the of these research a based on As in CL were for the systematic development of NLP changed the view of parsing as up the the systematic of of of into and so These also for in NLP such as of and so which are for treating the other hand, of CL to NLP did not this would in had to into more which required significant and on the NLP For example, we had to the into such as and so It is that, while the was to the hierarchy each stage in the was defined and related to each other through the single also that in the fields of computer changed was possible to in NLP. For example, the design of an abstract machine and its for in et al. systems for of and Tsujii Tsujii, and and so would be impossible in the broader fields of computer and much computational et al. the other hand, the in NLP. models in of the with the that we for the the NLP community with systematic ways of with of different models also which made systematic development of NLP systems possible. However, the in parsing still are the in of the number of was the sentence level, the That is, a sentence in which all are very Because of are (i.e., of or of in such as of and Tsujii using there are to a understanding is taken into This to the research language and and Motivation. I was interested in the topic of how to relate language with at the very beginning of my the my led to that a of could be used as a and was engaged in research of a system based on a and Tsujii However, resources such as a of processing of computer systems, and NLP technologies, such as were not available at the soon that the research would a whole of research topics in such as representation of common human ways of and so the topics had to deal with and of or the models that humans have may from one to I that the research target was through research in MT and parsing in the later of my career, I to that NLP research is if it how is in and that NLP are all related to of understanding and the same NLP as an engineering field, I it to be to have a definition of or information with which language is to be I would like to too much of research into and reasoning and to our research focus to the between language and As a research I to focus on the as the There were two for the was that colleagues at the two with which I was that, in to it had become increasingly for to of information in a of in diverse subject fields such as and In to the of they also had diverse that had to be with each In other they had a of shared by that was to be with information in other was that there were colleagues at the of who were interested in According to the on information in a by the group and research into at the of on between and and et al. the had been a natural of The was that information in a and were defined by the target and not by NLP had and to their own in the target Contributions. Although there had been a of research into information and for the there had been no to NLP to in the To address the at the of and the of a research in this this was a research we first had to tasks to to resources, and to not only NLP but also in the target the involvement of NLP and we found that a in the also began to be interested in research In to we a number of research in with colleagues the which led to of a at The first in with the The and shared It been and become one of the in NLP The research field of of NLP to is to et al. et al. and science et al. by the two include the et al. and Tsujii a of with their and Tsujii the et al. the et al. a design for information et al. an system based on Tsujii, and a system for et al. is one of the used in the To information in and how it was encoded in language, we not only from the of view but also from the of of and and and were to the et al. were with of the target which had been by the target to information in diverse in we developed a with which is used by the NLP In with we defined a set of NLP tasks et al. and Tsujii and Tsujii et al. and developed a set of et al. et al. et al. for which were to be into to of of et al. et al. a of this we between such as and and such as and and that (Figure The mapping between structures and the defined by was more complex than the mapping assumed by computational soon that, as an NLP information was very different from In particular, there were differences between the information that the to convey and in and the information that the readers (i.e., a group of to and from of the information that the to the would the information that they were interested of as an NLP made the mapping from language to information very different from the transfer phase in which to the same information in the source and target languages. While the of from structures (i.e., and their structures would only for the of and and these were to be with other The mapping to the transfer based on a bundle of the transfer in the level of representation, we first used the to climb up the hierarchy to the we called from which we to a set of to (Figure assumed that, based on surface of may be this would at a level of is, the same approach to simple transfer at the abstract Although this approach it soon reached its increasingly and we that this was of the nature of and to the approach based on a bundle of features (Figure et al. This to the research, which a language model In this information is assumed to be in the language The information is not in systems (Figure and et al. the other hand, our in the tasks and of In this it that structures more significant example, about an from different each As for the or of are fields such as and about an be made or with different of To the degree of of a we have to examine the of in which the is et al. and on the of an in a sentence, it may be as a and so The in which information by a be to and an research of an is the of in which an NLP system is used to a set of different from different to a of et al. In this a system must be to whether two in different be as the same in a To the system must be to the in which two by the of the This may also require structures to be taken into on the we forms of (i.e., by the target and diverse which include The of information in with these resources to research topics on the between language and of the target Because such as have views on which of information their knowledge, we the and of that may have research in the general with which NLP technologies such as parsing have been less than we research into the of information to the of is the nature of human The CL and NLP to reasoning as a of process based on However, the reasoning that the in the may not be so in example, we found that the reasoning by on is based on between For example, they that a A is to be in an by a that a which is to is in the same of The between A and is based on the between their Because such among are in their in language models on a of would be to their (i.e., a of also to such it may be to use of information, such as of collections of and so to such among and to reasoning based on have the progress and significant changes that models and have to the field of NLP. This is a in which in broader fields of computer up to and the NLP changes by and are and have had a not only on NLP but also and other of It is a the of diverse NLP I it the NLP by NLP with the processing of other information and with and so fields such as and reasoning based on used different and processing They the same basis of and It is much to forms of that NLP in and NLP with are more than we The research of the with which I are on these et al. et al. and the NLP based on language models is increasingly from other research disciplines that the study of language. The nature of and also the way of NLP systems the are very I that the may to by MT could the by computational but and systematic analysis of this to and the of the through analysis of parsing when the of analysis based on by other disciplines, and of NLP systems and the other hand, CL to language as a system or focus on study on aspects of that language in NLP systems, with NLP CL would be to the of their and to a theoretical basis for of NLP is the to NLP and was to the field of NLP by my who was a of the away It is that I could not my and with my career of almost 50 I have research into NLP at institutes Kyoto of the of and I the on of the research and who worked with at these my research could not have in the way that it I their
This work provides the first in-depth analysis of genre in Universal Dependencies (UD). In contrast to prior work on genre identification which uses small sets of well-defined labels in mono-/bilingual setups, UD contains 18 genres with varying degrees of specificity spread across 114 languages. As most treebanks are labeled with multiple genres while lacking annotations about which instances belong to which genre, we propose four methods for predicting instance-level genre using weak supervision from treebank metadata. The proposed methods recover instance-level genre better than competitive baselines as measured on a subset of UD with labeled instances and adhere better to the global expected distribution. Our analysis sheds light on prior work using UD genre metadata for treebank selection, finding that metadata alone are a noisy signal and must be disentangled within treebanks before it can be universally applied.
A linguistic norm (literary norm) is the rules for the use of speech means in a certain period of the development of the literary language, i.e. rules of pronunciation, word use, use of traditionally established grammatical, stylistic and other linguistic means adopted in social and linguistic practice. This is a uniform, exemplary, generally recognized use of language elements (words, phrases, sentences). Linguistic norms are a historical phenomenon. Changes in literary norms are due to the constant development of the language. What was the norm in the last century and even 15-20 years ago today can become a deviation from it. The norms help the literary language to maintain its integrity and comprehensibility. They protect the literary language from the flow of dialectal speech, social and professional jargon, and vernacular. This allows the literary language to fulfill its main function - cultural.
DEFI is a prototype computer tool aimed at ranking (from most to least relevant) the French translations of an English lexical item in context. This paper deals with the strategies used by DEFI to recognize multi-word units (mwus) in running text. Any lexical unit included in the lexical database used in the project (a merge of the Oxford/Hachette and Robert/Collins English-to-French dictionaries) and longer than a single word is submitted to a surface parser, and the same process is applied to the user ’s text. A program written in Prolog assesses the quality of the match between the parsed user’s text and candidate mwus retrieved from the project’s lexical database. The matcher is able to account for some of the distortions undergone by the mwu, e.g. movement of a constituent as a result of relativization or passivization.
Statistical learning (SL) refers to the ability to extract regularities in the environment and has been well-documented to play a key role in speech segmentation and language acquisition. Whether SL requires top-down attention is an unresolved question. The current study examined whether SL can occur outside the focus of attention. Participants either focused or diverted their attention away from a heard nonsense language. Visual attention was taxed by requiring tracking of multiple randomly moving dots. Linguistic attention was taxed through a self-paced reading task. SL was assessed with an explicit familiarity rating task, and an implicit reaction-time (RT) based memory task. Explicit learning was only reduced when linguistic resources were taxed, but unimpaired when visual resources were taxed. On our implicit measure of SL, learning was unimpaired when attention was taxed. These results suggest implicit aspects of SL to be more robust to an attention diversion as compared to explicit aspects of SL. Findings suggest L2 learners may engage in demanding visual tasks, as well as offer insight into the neurocognitive underpinnings of SL.
Unlike English letters, Chinese characters have rich and specific meanings. Usually, the meaning of a word can be derived from its constituent characters in some way. Several previous works on syntactic parsing propose to annotate shallow word-internal structures for better utilizing character-level information. This work proposes to model the deep internal structures of Chinese words as dependency trees with 11 labels for distinguishing syntactic relationships. First, based on newly compiled annotation guidelines, we manually annotate a word-internal structure treebank (WIST) consisting of over 30K multi-char words from Chinese Penn Treebank. To guarantee quality, each word is independently annotated by two annotators and inconsistencies are handled by a third senior annotator. Second, we present detailed and interesting analysis on WIST to reveal insights on Chinese word formation. Third, we propose word-internal structure parsing as a new task, and conduct benchmark experiments using a competitive dependency parser. Finally, we present two simple ways to encode word-internal structures, leading to promising gains on the sentence-level syntactic parsing task.
This is a small hand-annotated partial treebank of Modern Tibetan, primarily in CoNLL-U format. Some texts were POS-tagged by machine, and then dependency relations between verbs and their arguments were added by hand. Other texts include only dependency relations and relevant POS-tags. A number of the texts have English translations which have been manually aligned to the Tibetan text. This work was created as part of the AHRC-funded project <em>Lexicography in Motion</em> (PI Ulrich Pagel, 2017-2021).
In the prefaces to his French translations of Tacitus (1640) and Lucian (1654), Nicolas Perrot d’Ablancourt rationalizes his substantial revisions by appealing to the canons of French literary taste that his translations helped to form. This view resulted in translations that were clearer and more stylistically felicitous than the source texts, but also bowdlerized so as to avoid any violation of French linguistic norms as well as any moral offense. Perrot d’Ablancourt is very much aware that his discursive strategies flouted conventional notions of equivalence. Yet he makes clear that his domesticating choices are not arbitrary, but based on an interpretation that displays an acute sense of historical difference. He just does not feel that this difference is worth preserving in itself and certainly not at the cost of departing from an elegant style as he conceives it.
While the predictive performance of modern statistical dependency parsers relies heavily on the availability of expensive expert-annotated treebank data, not all annotations contribute equally to the training of the parsers. In this paper, we attempt to reduce the number of labeled examples needed to train a strong dependency parser using batch active learning (AL). In particular, we investigate whether enforcing diversity in the sampled batches, using determinantal point processes (DPPs), can improve over their diversity-agnostic counterparts. Simulation experiments on an English newswire corpus show that selecting diverse batches with DPPs is superior to strong selection strategies that do not enforce batch diversity, especially during the initial stages of the learning process. Additionally, our diversityaware strategy is robust under a corpus duplication setting, where diversity-agnostic sampling strategies exhibit significant degradation.
Different linearizations have been proposed to cast dependency parsing as\nsequence labeling and solve the task as: (i) a head selection problem, (ii)\nfinding a representation of the token arcs as bracket strings, or (iii)\nassociating partial transition sequences of a transition-based parser to words.\nYet, there is little understanding about how these linearizations behave in\nlow-resource setups. Here, we first study their data efficiency, simulating\ndata-restricted setups from a diverse set of rich-resource treebanks. Second,\nwe test whether such differences manifest in truly low-resource setups. The\nresults show that head selection encodings are more data-efficient and perform\nbetter in an ideal (gold) framework, but that such advantage greatly vanishes\nin favour of bracketing formats when the running setup resembles a real-world\nlow-resource configuration.\n
In this paper we address the problem of the combined representation of heterogeneous sources of knowledge within a unique and homogeneous data structure. The goal is ultimately to enable the holistic processing of linguistic and world knowledge, where all the dimensions may interact seamlessly. We focus here on bridging the gap between Syntax and Semantics. We propose to link an abstract grammar to an existing lexical network through the adoption of the same underlying graph structure. The resulting structure may be seen as a multi-layer linguistic network. The solution we introduce for the abstract grammar layer relies on a graph-theoretic interpretation of Property Grammar. The typed structure we propose supersedes both phrase structure and dependency structure, which are covered with specific relation types-Constituency and Dependency respectively. We present a procedure to derive the grammar from an annotated corpus, and we illustrate the procedure with the French Treebank.
There is empirical evidence that expected yet not current affect predicts decisions.However, common research designs in affective decision-making show consistent methodological problems (e.g., conceptualization of different emotion concepts; measuring only emotional valence, but not arousal).We developed a gambling task that systematically varied learning experience, average feedback balance and feedback consistency.In Experiment 1 we studied whether predecisional current affect or expected affect predict recurrent gambling responses.Furthermore, we exploratively examined how affective information is represented on a neuronal level in Experiment 2. Expected and current valence and arousal ratings as well as Blood Oxygen Level Dependent (BOLD) responses were analyzed using a within-subject design.We used a generalized mixed effect model to predict gambling responses with the different affect variables.Results suggest a guiding function of expected valence for decisions.In the anticipation period, we found activity in brain areas previously associated with valencegeneral processing (e.g., anterior cingulate cortex, nucleus accumbens, thalamus) mostly independent of contextual factors.These findings are discussed in the context of the idea of a valence-general affective work-space, a goal-directed account of emotions, and the hypothesis that current affect might be used to form expectations of future outcomes.In conclusion, expected valence seems to be the best predictor of recurrent decisions in gambling tasks.
Recent years have witnessed significant improvement in ASR systems to recognize spoken utterances. However, it is still a challenging task for noisy and out-of-domain data, where substitution and deletion errors are prevalent in the transcribed text. These errors significantly degrade the performance of downstream tasks. In this work, we propose a BERT-style language model, referred to as PhonemeBERT, that learns a joint language model with phoneme sequence and ASR transcript to learn phonetic-aware representations that are robust to ASR errors. We show that PhonemeBERT can be used on downstream tasks using phoneme sequences as additional features, and also in low-resource setup where we only have ASR-transcripts for the downstream tasks with no phoneme information available. We evaluate our approach extensively by generating noisy data for three benchmark datasets - Stanford Sentiment Treebank, TREC and ATIS for sentiment, question and intent classification tasks respectively. The results of the proposed approach beats the state-of-the-art baselines comprehensively on each dataset.
Neural machine translation (NMT) models are typically trained using a softmax cross-entropy loss where the softmax distribution is compared against the gold labels. In low-resource scenarios and NMT models tend to perform poorly because the model training quickly converges to a point where the softmax distribution computed using logits approaches the gold label distribution. Although label smoothing is a well-known solution to address this issue and we further propose to divide the logits by a temperature coefficient greater than one and forcing the softmax distribution to be smoother during training. This makes it harder for the model to quickly over-fit. In our experiments on 11 language pairs in the low-resource Asian Language Treebank dataset and we observed significant improvements in translation quality. Our analysis focuses on finding the right balance of label smoothing and softmax tempering which indicates that they are orthogonal methods. Finally and a study of softmax entropies and gradients reveal the impact of our method on the internal behavior of our NMT models.
The article analyzes transformation of forms of degrees of comparison of adjectives in live television broadcasting. Particular attention is paid to the specific properties of different forms of degrees of comparison of adjectives. To analyze the peculiarities of their use for errors in speech of television journalists, associated with non-compliance with linguistic norms on ways to avoid these errors, to make appropriate recommendations to television journalists. The main method we use is to observe the speech of live TV journalist, we used during the study methods of comparative analysis of comparison of theoretical positions from the work of individual linguists and journalism sat down as well as texts that sounded in the speech of journalists. Our objective is to trace these transformations and develop a certain attitude towards them in our researches of the language of the media and practicing journalists to support positive trends in the development of the broadcasting on TV and give recommendations for overcoming certain negative trends. Improving the live broadcasting of television journalists, in particular the work on deepening the language skills will contribute to the modernization of some trends in the reasonable expediency of the transformation of certain phenomena, modernization of some tendencies concerning the reasonable expedient transformation of separate grammatical phenomena and categories and at braking and in general stopping of processes of transformation of negative unreasonable not expedient. This fully applies primarily to attempts to transform the forms of degrees of comparison of adjectives and this explains importance of the results achieved in these study.
Recent advancements in language models based on recurrent neural networks and transformers architecture have achieved state-of-the-art results on a wide range of natural language processing tasks such as pos tagging, named entity recognition, and text classification. However, most of these language models are pre-trained in high resource languages like English, German, Spanish. Multi-lingual language models include Indian languages like Hindi, Telugu, Bengali in their training corpus, but they often fail to represent the linguistic features of these languages as they are not the primary language of the study. We introduce HinFlair, which is a language representation model (contextual string embeddings) pre-trained on a large monolingual Hindi corpus. Experiments were conducted on 6 text classification datasets and a Hindi dependency treebank to analyze the performance of these contextualized string embeddings for the Hindi language. Results show that HinFlair outperforms previous state-of-the-art publicly available pre-trained embeddings for downstream tasks like text classification and pos tagging. Also, HinFlair when combined with FastText embeddings outperforms many transformers-based language models trained particularly for the Hindi language.
Constructive interactions through discussion forums allow students to open their horizons and thought processes to acquire more knowledge and develop skills. Thus, discussion forums play an important role in supporting learning. Additionally, the discussion forum provides the content for creating a knowledge repository. It contains discussion threads related to key course topics that are debated by the students. One approach to understanding the student learning experience is through the analysis of the discussion threads. This research proposes the application of discourse analysis and collaborative learning frameworks to discussion forums to gain further insights into the student's learning in a classroom. It is a foray into discourse analysis using in-class discussions. It demonstrates the application of Soller's framework and Penn Discourse Treebank (PDTB) to understand interactions at the discourse and semantic level. It also shows the use of unsupervised automated techniques to diagnose interactions in textual data. In this paper, we present an Integrated Discourse Analysis and Collaborative Learning Skills (IDALS) framework based on in-class discussions. We describe our experiences of applying IDALS framework and evaluating the solution model in a graduate in-class discussion forum. We also highlight the benefits of using visualizations to present the insights to the instructors.
Machine learning training methods depend plentifully and intricately on hyperparameters, motivating automated strategies for their optimisation. Many existing algorithms restart training for each new hyperparameter choice, at considerable computational cost. Some hypergradient-based one-pass methods exist, but these either cannot be applied to arbitrary optimiser hyperparameters (such as learning rates and momenta) or take several times longer to train than their base models. We extend these existing methods to develop an approximate hypergradient-based hyperparameter optimiser which is applicable to any continuous hyperparameter appearing in a differentiable model weight update, yet requires only one training episode, with no restarts. We also provide a motivating argument for convergence to the true hypergradient, and perform tractable gradient-based optimisation of independent learning rates for each model parameter. Our method performs competitively from varied random hyperparameter initialisations on several UCI datasets and Fashion-MNIST (using a one-layer MLP), Penn Treebank (using an LSTM) and CIFAR-10 (using a ResNet-18), in time only 2-3x greater than vanilla training.
In this paper, we address the representation of coordinate constructions in\nEnhanced Universal Dependencies (UD), where relevant dependency links are\npropagated from conjunction heads to other conjuncts. English treebanks for\nenhanced UD have been created from gold basic dependencies using a heuristic\nrule-based converter, which propagates only core arguments. With the aim of\ndetermining which set of links should be propagated from a semantic\nperspective, we create a large-scale dataset of manually edited syntax graphs.\nWe identify several systematic errors in the original data, and propose to also\npropagate adjuncts. We observe high inter-annotator agreement for this semantic\nannotation task. Using our new manually verified dataset, we perform the first\nprincipled comparison of rule-based and (partially novel) machine-learning\nbased methods for conjunction propagation for English. We show that learning\npropagation rules is more effective than hand-designing heuristic rules. When\nusing automatic parses, our neural graph-parser based edge predictor\noutperforms the currently predominant pipelinesusing a basic-layer tree parser\nplus converters.\n
The purpose of this study is to examine the orthographic and phonological characteristics of the Yeongsan Sillok(the biography of Yeongsan), published in Jeollabuk-do in the early 20th century. The author of this book is considered to be Jang Bong-seon, an educator from Jeongeup city in Jeollabuk-do. Accordingly, it is expected that this book contains the orthographic characteristics and attitudes toward the language of young intellectuals in Jeollabuk-do in the early 20th century. In Chapter 3, we looked at the orthographic characteristics of this book. The writing characteristics of this book largely follow the characteristics of the 19th century Jeollabuk-do dialect based on the tradition of modern Korean. However, a transitional characteristic of the language transforming into present-day Korean was also present. Although only a few examples have been confirmed, the writing of double consonant letters for tense consonant are gradually similar to the notation method of modern Korean. This can be understood as a dissolution process. At the same time, with the exception of some circumstances of verbs, the tendency to split consonants is widely confirmed, and the modern Korean notation for the /ㄹㄹ/ chain (ㄹㄴ, ​​ㄹㅇ) is gradually changing to ㄹㄹ. Above all, the fact that the notation of ․ or diphthong after sibilants no longer appears in this book is a characteristic feature that differs from data from the Jeollabuk-do region of the same period. This writing trend seems to be related to a set of linguistic norms compiled in the first half of the 20th century. Recalling that the author of this book established a private school in the 1920s and 1930s and devoted himself to educational activities, this assumption is somewhat probable. In Chapter 4, we looked at the phonological characteristics of the Yeongsan Sillok(the biography of Yeongsan). Front-vowelization was very active inside the morpheme, but at the morpheme boundary, it appeared only in the environment behind c. The simple vowelization of jə>e is confirmed throughout the interior and boundary of the morpheme, and it must have been a productive phonological phenomenon in the Jeollabuk-do dialect in the early 20th century, as hypercorrection types also appeared. Regarding the alternation of the ending ‘-a/ə’, when the stem vowel is ‘ø’, there is a high tendency to combine these to ‘-ə’. This is different from the 19th century and modern Jeollabuk-do dialects. In the case of umlauts, only very limited examples were shown. And although t-palatalization is quite actively realized, only a few examples of k-palatalization were shown. Through this realization of phonological phenomena, we were able to confirm whether the young intellectuals in the Jeollabuk-do region in the early 20th century had linguistic attitudes toward the Jeollabuk-do dialect. In this book, the typical phonological phenomenon of the Jeollabuk-do dialect was confirmed only to a very limited extent due to its negative evaluation by the author.
The main motivation behind this exam document is to look at the extent to which EWOM among customers can affect the brand image and the intent of buying the consumer in the clothing industry. A key condition display process is linked to the E-WOM impacts survey on brand image and buyer's purchase target. The exploration program was tested using an example of 385 respondents who included information within online purchasing groups and examined buyers of Pakistan's textile industry at the time of the investigation. The document recalls the methodologies to help a brand profitably through client-based social networking on the web, as well as typical suggestions for delegated websites and dialogues to enhance this note on a major path with people in their online dating. This explorative document extends the winning image rating to another set, in particular e-WOM. This document provides profitable knowledge on e-WOM estimation, brand image and purchasing expectations of the purchaser in the clothing industry and provides a facility for future search for tagging items.
In this paper, we propose a method for learning representations in the space of Gaussian-like distribution defined on a novel geometrical space called Kinematic space. The utility of non-Euclidean geometry for deep representation learning has recently been in vogue, specifically models of hyperbolic geometry such as Poincaré and Lorentz models have proven useful for learning hierarchical representations. Going beyond manifolds with constant curvature, albeit has better representation capacity might lead to unhanding of computationally tractable tools like Riemannian optimization methods. Here, we explore a pseudo-Riemannian auxiliary Lorentzian space called Kinematic space and provide a principled approach for constructing a Gaussian-like distribution, which is compatible with gradient-based learning methods, to formulate a probabilistic word embedding framework. Contrary to, mapping lexically distributed representations to a single point vector in Euclidean space, we advocate for mapping entities to density-based representations, as it provides explicit control over the uncertainty in representations. We test our framework by embedding WordNet-Noun hierarchy, a large lexical database, our experiments report strong consistent improvements in Mean Rank and Mean Average Precision (MAP) values compared to probabilistic word embedding frameworks defined on Euclidean and hyperbolic spaces. We show an average improvement of 72.68% in MAP and 82.60% in Rank compared to the hyperbolic version. Our work serves as evidence for the utility of novel geometrical spaces for learning hierarchical representations.
Cloud-based enterprise search services (e.g., AWS Kendra) have been\nentrancing big data owners by offering convenient and real-time search\nsolutions to them. However, the problem is that individuals and organizations\npossessing confidential big data are hesitant to embrace such services due to\nvalid data privacy concerns. In addition, to offer an intelligent search, these\nservices access the user search history that further jeopardizes his/her\nprivacy. To overcome the privacy problem, the main idea of this research is to\nseparate the intelligence aspect of the search from its pattern matching\naspect. According to this idea, the search intelligence is provided by an\non-premises edge tier and the shared cloud tier only serves as an exhaustive\npattern matching search utility. We propose Smartness At Edge (SAED mechanism\nthat offers intelligence in the form of semantic and personalized search at the\nedge tier while maintaining privacy of the search on the cloud tier. At the\nedge tier, SAED uses a knowledge-based lexical database to expand the query and\ncover its semantics. SAED personalizes the search via an RNN model that can\nlearn the user interest. A word embedding model is used to retrieve documents\nbased on their semantic relevance to the search query. SAED is generic and can\nbe plugged into existing enterprise search systems and enable them to offer\nintelligent and privacy-preserving search without enforcing any change on them.\nEvaluation results on two enterprise search systems under real settings and\nverified by human users demonstrate that SAED can improve the relevancy of the\nretrieved results by on average 24% for plain-text and 75% for encrypted\ngeneric datasets.\n
Literary Works byAkaki Tsereteli are considered as versatile and diverse. In his works he touches upon almost everything by his poetry, prose, journalism or public work. It is obvious that he established "a type of versatile writer who is equally engaged in prose, poetry, journalism, dramaturgy, translations, children's literature and fables”. He was an extremely optimistic person who deeply believed in the future. The following words from one of his works seem amazingly and expressive: “Even if you kill a swallow, Spring will definitely come”. Connection between the old and the new forms, that is clearly shown within this emotionally colored expression, has become the goal of the research. We tried to find an answer to the question- what is the role of using old Georgian forms in Akaki's work?! Given paper analyses the samples such as: 1. Using proper name by its stem form in nominative case; 2. Ending words by - მან [-man] in the ergative form; 3. Full stems of demonstrative pronouns - ‘ამ’ [am], ‘ეგ’ [eg] (=this, that); 4. Using postposition – ‘ზე’ [ze] (=on), along with the forms - ზედ [-zed] and -ზედა [-zeda] (=on, over); 5. instrumental case forms formed by a suffix - ით [-it] (=with) (without postpositions); 6. Postposition and full agreement of attribute and antecedent 7. Characteristics of using inflection as a reflection of Old Georgian (გწყალობდესთ [gtskalobdet]...; გამოვჰკითხავ [gamovhkitkhav]...; ჰსვამ [hvsvam]...; ჰნიშნავს [hnishnavs]...; წარმოსთქვა [tsarmostkva]...; გასტეხე [gastekhe]...); 8. Using conjunction - ვით [vit] (=as/like) for comparison and so on. If we ask questions concerning the function of old Georgian forms in Akaki Tsereteli’s works, it becomes clear that they can be used for: 1. rhythm, emotiveness and expressiveness; 2. Preserving traditional forms, to maintain the connection between old and new Georgian. It should be mentioned that similar forms are equally reflected in Akaki’s prose and poetry which further reinforces the idea in favor of showing the connection between the old and the new and the desire to maintain this connection and always remember where we come from and who we are....This fact does not completely contradict the idea that Akaki is a representative of the generation that courageously rejected the old linguistic norms and contributed to the democratization (rapprochement process with the spoken language) of the literary language.
With a pair of oppositely valenced stimuli, rating the first one sometimes leads to a more extreme evaluation for the second (e.g., if the second is negatively valenced, rating the first stimulus would lead to a more negative rating for the second). We considered an evaluation bias in the case of clinical diagnosis relating to eating disorders. A population sample which included experienced clinical psychologists and psychiatrists showed partial evidence of an evaluation bias, when judging descriptions of individuals designed to be consistent with eating disorders or not. Quantum probability theory, the probability rules from quantum mechanics without any of the physics, is particularly well-suited to modeling the evaluation bias (and constructive influences generally), because a measurement (or judgment) can change the state of the system. We applied a previous quantum model to the present result, an extension of the model embodying noisy processes, and belief adjustment model. We discuss how model fits inform an examination of rationality in the observed behavior.
This paper investigates updates of Universal Dependencies (UD) treebanks in 23 languages and their impact on a downstream application. Numerous people are involved in updating UD's annotation guidelines and treebanks in various languages. However, it is not easy to verify whether the updated resources maintain universality with other language resources. Thus, validity and consistency of multilingual corpora should be tested through application tasks involving syntactic structures with PoS tags, dependency labels, and universal features. We apply the syntactic parsers trained on UD treebanks from multiple versions (2.0 to 2.7) to a clause-level sentiment extractor. We then analyze the relationships between attachment scores of dependency parsers and performance in application tasks. For future UD developments, we show examples of outputs that differ depending on version.