Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
Introduction<br><br> Penn Discourse Treebank Version 2.0 - German Translation was developed at the University of Potsdam's Applied Computational Linguistics group and consists of approximately one million tokens derived from Penn Discourse Treebank Version 2.0 (LDC2008T05). This data was translated into German and annotated for shallow discourse relations in the financial news domain.<br><br> The aim of the University of Pennsylvania's Penn Discourse Treebank (PDTB) project is to annotate the Wall Street Journal text in Treebank-2 with discourse relations. PDTB 2.0 contains 40,600 tokens of annotation relations. PDTB2-German is based on a subset of PDTB2.0 used in the 2016 CoNLL Shared Task on Multilingual Shallow Discourse Parsing. Data<br><br> Data is in CoNLL format. Text was automatically translated into German with deepL, and projections of the annotations using word alignments were produced with GIZA++. See the included documentation for more information on the relation annotations.<br><br> Source text and CoNLL format annotations are each presented in their own tab separated plain text file, encoded in UTF-8. Samples<br><br> Please view this source sample and annotation sample. Updates<br><br> None at this time. Copyright Portions © 1987-1989 Dow Jones & Company, Inc., © 2008, 2012, 2021 The Penn Discourse Treebank Group, © 2021 Manfred Stede, © 1993-1995, 2008, 2012, 2021 Trustees of the University of Pennsylvania
This article deals with a vision for the achievement of the project of a linguistic atlas of the dialects of Algeria supported by colored digital maps, showing the dialectical diversity of the selected region. And the circulation, the dialects used and current on the tongue of the inhabitants of Algeria We will focus in our research on the aspect of processing linguistic data in the manufacture of a digital linguistic atlas, in an attempt to invest computer data in describing local dialects and access to digital content, with the possibility of this content audio recordings of dialect variations, by providing our data bank through this linguistic research Also to pay tribute to the importance of the linguistic atlas in meeting the need of dialect workers for linguistic maps that identify the locations, nature and types of dialectal diversity in Algeria, The research also comes mainly to embody the digital principle and support the Arabic language by storing it digitally with the possibility of managing and printing linguistic maps according to the changes in them.
The Mongolian written language and the traditional Mongolian script were a “pre-modern” language and script that transcended ethnicity and dialect. The Mongolian script can be read in any dialect. However, modern languages demand pronunciation norms, making the Mongolian script unsuitable as the official script of a modern nation-state. The countries and regions that used the Mongolian script changed the script in the first half of the 20th century, except for the Mongolian ethnic areas of China which continued to use the original Mongolian script. This was possible as Mongolian was a minority language in China, and the scope of its use was limited. However, Inner Mongolia, too, faced the issue of the written language not conforming with the spoken language. Therefore, in the 1930s, Mongolian literary figures and others in Manchukuo attempted to unify the Mongolian written and spoken language based on the genbun itchi movement in Japan. Genbun itchi sought to unify the Japanese written and spoken language, and was translated literally as “üge üsüg-i nigen bolcaqui” in Mongolian. This effort was only partially successful in changing the Mongolian script to match the spoken pronunciation, and systematic genbun itchi could not be achieved. Since 1945, Inner Mongolia learned from the example of the Mongolian People’s Republic and transformed the style of Mongolian script to that of modern Mongolian while still using the Mongolian script. However, the discord between the Mongolian script and the spoken pronunciation remains unresolved, and to address this, changes are being made to the Mongolian script to this day as they have been for the last eight decades. It is important to know that the Mongolian script is not only the modern script used in Chinese territory, but also the traditional script of the written language common to areas using the Mongolian written language. For this reason, the Mongolian script must be passed on to future generations. A means for unifying the Mongolian language while using the Mongolian script is to follow the Hanyu Pinyin system, which has succeeded in standardizing pronunciation while using Chinese characters—namely, enhance the Inner Mongolian “standard-sounding” writing system so that it can also function at the sentence level. The orthography of the Cyrillic alphabet of Mongolia offers an objectively good example for the development of a standard-sounding writing system in Inner Mongolia. After all, the problem of the Mongolian script reform comes down to that of genbun itchi.
Dependency Parsing (DP) is the existence of sub-term/upper-term relations between the words that make up that sentence for each sentence in the text. DP serves to produce meaningful information for high-level applications. Correct labeling of the text corpus used in DP studies is very important. There will be mistakes in the results of the studies that will be performed with the wrongly-labeled text corpus. If text corpus is labeled manually or automatically by human beings, then faulty cases will occur. As a result of the cases that may arise from human factors or annotations used for labeling, faulty labels will be on treebanks. In order to prevent these errors, detection, and correction of possible faulty labeling is very important in terms of increasing the accuracy of the studies to be carried out. Manual correction of possible faulty labels requires great effort and time. The purpose of this study is to create a model that automatically finds possible faulty labels and offers new label suggestions for faulty labels. With the help of the proposed model, it is aimed to detect and correct possible faulty labels that are included in a text corpus, and to increase consistency among the text corpus of the same language. With the help of the developed model, suggesting new labels for faulty labels by a language expert will be a great convenient for the specialist. Another advantage of the model is that the developed model provides a language-independent structure. It has succeeded in obtaining successful results in finding and correcting potentially faulty labels in experimental studies for Turkish. An increase in accuracy has been detected in studies carried out for languages other than Turkish. In investigating the accuracy of the results obtained by the system, the results were analyzed with the help of 10 different language experts.
Mark Anderson, Anders Søgaard, Carlos Gómez-Rodríguez. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers). 2021.
Dependency parsing is an important task for Natural Language Processing (NLP). However, a mature parser requires a large treebank for training, which is still extremely costly to create. Tibetan is a kind of extremely low-resource language for NLP, there is no available Tibetan dependency treebank, which is currently obtained by manual annotation. Furthermore, there are few related kinds of research on the construction of treebank. We propose a novel method of multi-level chunk-based syntactic parsing to complete constituent-to-dependency treebank conversion for Tibetan under scarce conditions. Our method mines more dependencies of Tibetan sentences, builds a high-quality Tibetan dependency tree corpus, and makes fuller use of the inherent laws of the language itself. We train the dependency parsing models on the dependency treebank obtained by the preliminary transformation. The model achieves 86.5% accuracy, 96% LAS, and 97.85% UAS, which exceeds the optimal results of existing conversion methods. The experimental results show that our method has the potential to use a low-resource setting, which means we not only solve the problem of scarce Tibetan dependency treebank but also avoid needless manual annotation. The method embodies the regularity of strong knowledge-guided linguistic analysis methods, which is of great significance to promote the research of Tibetan information processing.
Previous studies have shown that dependency distance and its probability distribution can be applied as syntactic indicators of English as interlanguage. However, the universal application of these indicators has not been verified from the perspective of language typology. The issues are addressed in the present study based on a treebank of Chinese interlanguage of English and Japanese native speakers. The findings are as follows: (1) with the improvement of L2 proficiency, the MDDs of learners with different native language backgrounds gradually approach that of the target language in different patterns, and dependency distance is of universal significance as a metric to measure the development of interlanguage’s syntactic complexity; (2) Chinese interlanguage also follows the principle of least effort, and its probability distribution of dependency distance, like those of natural languages, presents a power–law distribution, which can successfully fit the Zipf-Alekseev distribution; (3) the right truncated modified Zipf-Alekseev distribution can be used to measure Chinese interlanguage proficiency, and the fitting parameters of the probability distribution of dependency distance as a metric of interlanguage proficiency are also of universal value.
In 2021, we organized the second iteration of a shared task dedicated to the underlying units used in discourse parsing across formalisms: the DISRPT Shared Task (Discourse Relation Parsing and Treebanking). Adding to the 2019 tasks on Elementary Discourse Unit Segmentation and Connective Detection, this iteration of the Shared Task included for the first time a track on discourse relation classification across three formalisms: RST, SDRT, and PDTB. In this paper we review the data included in the Shared Task, which covers nearly 3 million manually annotated tokens from 16 datasets in 11 languages, survey and compare submitted systems and report on system performance on each task for both annotated and plain-tokenized versions of the data.
Swift detection of faces with emotional meaning underlies fruitful social relationships. Although previous studies using a visual search paradigm have demonstrated rapid detection of emotional facial expressions, whether it is attributable to emotional/motivational significance remains to be clarified. We examined this issue by excluding the influence of visual factors on the rapid detection of faces with emotional meaning. First, participants were engaged in an associative learning task wherein neutral faces were associated with either monetary rewards, monetary punishments, or zero outcome in order for the neutral faces to acquire positive, negative, and no emotional value, respectively. Then, during the visual search task, the participants detected a target-neutral face associated with high reward or punishment from among newly presented neutral faces. In Experiment 1, neutral faces associated with high reward and punishment values were more rapidly detected than those without monetary outcomes. In Experiment 2, highly rewarded and highly punished neutral faces were more rapidly detected than neutral faces associated with low monetary reward/punishment. Analyses of ratings confirmed that the learned neutral faces acquired emotional value, and the reaction times were negatively related to arousal ratings. These results suggest that the emotional/motivational significance promotes the rapid detection of emotional faces.
As in all other fields, linguistics is accelerating the process of adapting to digital technologies. Consequently, it is important to process the traditional linguistic norms of natural language for computer programs and information systems. One such important task in NLP is tagging parts of speech. Part-of-speech tagging (abbreviation: (POS tagging or PoS tagging or POST) in Russian “частеречная разметка”) is a stage of automatic text processing, the function of it which is a series of words (forms) used in the text and it is to determine grammatical features. With this function, POS-tagging is one of the first steps in automatic text analysis. The article discusses the need for tagging parts of speech, tagging methods.
Music’s ability to influence exercise performance is well known, but the converse, how exercise influences music listening, remains largely unknown. Exercise can elevate arousal, positive affect, and neurotransmitters including dopamine, which are involved in musical pleasure. Here we examine how exercise influences music enjoyment, and test for a modulatory role of arousal, affect, and dopamine. Before and after exercise (12 min of vigorous running) and a rest control session, participants ( N = 20) listened to music clips and rated their enjoyment and subjective arousal; we also collected ratings of affect and eye-blink rates, an established predictor of dopamine activity. Ratings of musical enjoyment increased significantly after exercise, but not after the rest control condition. While changes in subjective arousal ratings did not differ between exercise and control conditions, change in subjective arousal correlated with change in music enjoyment. After exercise, the change in music enjoyment had a positive but non-significant correlation with change in eye-blink rates ( r =.36). Positive affect increased more after exercise than after the control session, but the change in positive affect did not correlate with change in music enjoyment. In sum, exercise leads to increased musical enjoyment, and this effect was related to changes in arousal.
Existing family language policy (FLP) scholarship has been criticised for insufficiently addressing children’s voices and perspectives on their multilingual experiences, as well as lacking representation and heterogeneity in terms of studies involving multilingual families from diverse family types, languages, and contexts outside the experiences of Western middle-class bilingual families. Against this backdrop, this paper examines the multilingual familial experiences of three Ethiopian and Eritrean migrant families in Sweden by paying particular attention to children’s agency and caregiver-children dynamics in FLP making. The study draws on multimodal biographic data obtained from children and parents through language portrait methods of body and space mapping activities, post-mapping narration, and semi-structured interviews. The data are analysed in light of Smith-Christmas’s (2020) framework, which views child agency in FLP at the intersection of compliance regime, linguistic competence, linguistic norms, and power dynamics. The findings reveal that the process of FLP making is characterised as a process that is (1) filled with language choice dilemmas triggered by competing linguistic demands, (2) in part shaped by the family constellation via power dynamics between family members, and (3) mediated by family members’ varied linguistic proficiencies in majority and minority languages. Additionally, children’s agency about which language they choose to use impacts the language practices of the home, as they tend to establish their own linguistic norms within the home by overrunning the negotiated language policy set by caregivers.
Sentiment analysis is one of the prominent research areas in data mining and knowledge discovery, which has proven to be an effective technique for monitoring public opinion. The big data era with a high volume of data generated by a variety of sources has provided enhanced opportunities for utilizing sentiment analysis in various domains. In order to take best advantage of the high volume of data for accurate sentiment analysis, it is essential to clean the data before the analysis, as irrelevant or redundant data will hinder extracting valuable information. In this paper, we propose a hybrid feature selection algorithm to improve the performance of sentiment analysis tasks. Our proposed sentiment analysis approach builds a binary classification model based on two feature selection techniques: an entropy-based metric and an evolutionary algorithm. We have performed comprehensive experiments in two different domains using a benchmark dataset, Stanford Sentiment Treebank, and a real-world dataset we have created based on World Health Organization (WHO) public speeches regarding COVID-19. The proposed feature selection model is shown to achieve significant performance improvements in both datasets, increasing classification accuracy for all utilized machine learning and text representation technique combinations. Moreover, it achieves over 70% reduction in feature size, which provides efficiency in computation time and space.
Adapting threat-related memories towards changing environments is a fundamental ability of organisms. One central process of fear reduction is suggested to be extinction learning, experimentally modeled by extinction training that is repeated exposure to a previously conditioned stimulus (CS) without providing the expected negative consequence (unconditioned stimulus, US). Although extinction training is well investigated, evidence regarding process-related changes in neural activation over time is still missing. Using optimized delayed extinction training in a multicentric trial we tested whether: 1) extinction training elicited decreasing CS-specific neural activation and subjective ratings, 2) extinguished conditioned fear would return after presentation of the US (reinstatement), and 3) results are comparable across different assessment sites and repeated measures. We included 100 healthy subjects (measured twice, 13-week-interval) from six sites. 24 h after fear acquisition training, extinction training, including a reinstatement test, was applied during fMRI. Alongside, participants had to rate subjective US-expectancy, arousal and valence. In the course of the extinction training, we found decreasing neural activation in the insula and cingulate cortex as well as decreasing US-expectancy, arousal and negative valence towards CS+. Re-exposure to the US after extinction training was associated with a temporary increase in neural activation in the anterior cingulate cortex (exploratory analysis) and changes in US-expectancy and arousal ratings. While ICCs-values were low, findings from small groups suggest highly consistent effects across time-points and sites. Therefore, this delayed extinction fMRI-paradigm provides a solid basis for the investigation of differences in neural fear-related mechanisms as a function of anxiety-pathology and exposure-based treatment.
Chat-based counselling has become increasingly popular in the era of telecommunication. The need for accessible therapy has been exacerbated by the COVID-19 pandemic. Given its text-based nature, chat-based counselling provides an opportunity for machine-based analysis. It even has the potential to provide machine-based counselling services. However, the informational resources for machine-based analysis and interaction are rather scarce especially in a Japanese-language context. We created a Japanese dictionary for sentiment analysis, using a technique via machine-based text analysis, tailored for counselling related text. It includes 2389 words that were frequently used in chat-based counselling corpora. The following attributes were included for each word: (1) valence rating by the general public, (2) valence rating by clinical psychologists, (3) emotionality, and (4) body-relatedness.
The new and growing field of Quantitative Dependency Syntax has emerged at the crossroads between Dependency Syntax and Quantitative Linguistics. One of the main concerns in this field is the statistical patterns of syntactic dependency structures. These structures, grouped in treebanks, are the source for statistical analyses in these and related areas; dozens of scores devised over the years are the tools of a new industry to search for patterns and perform other sorts of analyses. The plethora of such metrics and their increasing complexity require sharing the source code of the programs used to perform such analyses. However, such code is not often shared with the scientific community or is tested following unknown standards. Here we present a new open-source tool, the Linear Arrangement Library (LAL), which caters to the needs of, especially, inexperienced programmers. This tool enables the calculation of these metrics on single syntactic dependency structures, treebanks, and collection of treebanks, grounded on ease of use and yet with great flexibility. LAL has been designed to be efficient, easy to use (while satisfying the needs of all levels of programming expertise), reliable (thanks to thorough testing), and to unite research from different traditions, geographic areas, and research fields.
Text discourse parsing weighs importantly in understanding information flow and argumentative structure in natural language, making it beneficial for downstream tasks. While previous work significantly improves the performance of RST discourse parsing, they are not readily applicable to practical use cases: (1) EDU segmentation is not integrated into most existing tree parsing frameworks, thus it is not straightforward to apply such models on newly-coming data. (2) Most parsers cannot be used in multilingual scenarios, because they are developed only in English. (3) Parsers trained from single-domain treebanks do not generalize well on out-of-domain inputs. In this work, we propose a document-level multilingual RST discourse parsing framework, which conducts EDU segmentation and discourse tree parsing jointly. Moreover, we propose a cross-translation augmentation strategy to enable the framework to support multilingual parsing and improve its domain generality. Experimental results show that our model achieves state-of-the-art performance on document-level multilingual RST parsing in all sub-tasks.
The study aimed to investigate if hentai consumers differed from other pornography consumers regarding their attachment style, attraction to, and desire for romantic relationships with anime characters and humans. Pornography consumers were categorized into three groups. The first group consumed both hentai and human pornography (hentai consumers), the second consumed human pornography but not hentai (non-hentai), and the third did not consume hentai or human pornography (non-porn). Two hundred and eight participants completed an online study that involved self-report surveys and an image rating task. The results revealed that hentai consumers did not differ from non-hentai or non-porn consumers on avoidant attachment. However, among females, hentai consumers were higher on anxious attachment compared to non-porn consumers. For the image rating task, hentai consumers rated anime characters more attractive than non-hentai and non-porn consumers. However, there were no group differences for the image ratings of real people. Hentai consumers indicated stronger romantic desire towards anime characters compared to non-hentai and non-porn consumers; there were no group differences in romantic desire for humans. The findings highlight the importance of differentiating individuals who consume hentai and those who do not.
In this paper, the process of creating a Dependency Treebank for tweetsin Urdu,a morphologically rich and less-resourced languageis described. The 500 Urdu tweets treebank iscreated by manually annotating the treebank withlemma, POS tags, morphological and syntacticrelations using the Universal Dependencies annotation scheme, adopted to the peculiarities of Urdu social media text. annotation process is evaluated through Inter-annotator agreement for dependency relations and total agreement of 94.5% and resultant weighted Kappa = 0.876was observed. The treebank is evaluated through 10-fold cross validation using Maltparserwith various feature settings. Results show average UAS score of 74%, LAS score of 62.9% and LA score of 69.8%.
ABSTRACT: Pain-related learning mechanisms likely play a key role in the development and maintenance of chronic pain. Previous smaller-scale studies have suggested impaired pain-related learning in patients with chronic pain, but results are mixed, and chronic back pain (CBP) particularly has been poorly studied. In a differential conditioning paradigm with painful heat as unconditioned stimuli, we examined pain-related acquisition and extinction learning in 62 patients with CBP and 61 pain-free healthy male and female volunteers using valence and contingency ratings and skin conductance responses. Valence ratings indicate significantly reduced threat and safety learning in patients with CBP, whereas no significant differences were observed in contingency awareness and physiological responding. Moreover, threat learning in this group was more impaired the longer patients had been in pain. State anxiety was linked to increased safety learning in healthy volunteers but enhanced threat learning in the patient group. Our findings corroborate previous evidence of altered pain-related threat and safety learning in patients with chronic pain. Longitudinal studies exploring pain-related learning in (sub)acute and chronic pain are needed to further unravel the role of aberrant pain-related learning in the development and maintenance of chronic pain.
Homomorphic encryption (HE) and garbled circuit (GC) provide the protection for users' privacy. However, simply mixing the HE and GC in RNN models suffer from long inference latency due to slow activation functions. In this paper, we present a novel hybrid structure of HE and GC gated recurrent unit (GRU) network, CRYPTOGRU, for low-latency secure inferences. CRYPTOGRU replaces computationally expensive GC-based tanh with fast GC-based ReLU, and then quantizes sigmoid and ReLU to smaller bit-length to accelerate activations in a GRU. We evaluate CRYP-TOGRU with multiple GRU models trained on 4 public datasets. Experimental results show CRYPTOGRU achieves top-notch accuracy and improves the secure inference latency by up to 138 over one of the state-of-the-art secure networks on the Penn Treebank dataset.
Research has shown that patients with a social anxiety disorder (SAD) show social performance deficits. These deficits are a maintaining factor in SAD, as mending social behavior improves interpersonal judgments and reduces social anxiety. Thus finding ways to enhance social behavior is evidently of importance in the treatment of SAD. This double-blind, placebo-controlled study investigated the effect of an intranasal administration of the hormone oxytocin (24 IU) on social behavior and anxious appearance in SAD patients (N = 40) and healthy controls (N = 39). Forty minutes after oxytocin administration participants were submitted to two live social situations (i.e., a waiting room situation and a getting acquainted task). The participants ('self-rated') and observers ('observer-rated') scored participants' social behavior and anxious appearance. Participants also rated their positive and negative affect. Confirming the social performance deficits in SAD, observers regarded SAD patients as more anxious and less socially skilled than healthy controls. Results indicated oxytocin-induced improvement of observer-rated social behavior in SAD patients compared to placebo but only in the getting acquainted task. This effect was not perceived as such by patients themselves and did not improve their affect ratings. In conclusion, this study found support for the idea that oxytocin helps SAD patients to perform better in social interactions, although this improvement seemed context-dependent (i.e., only present in the getting-acquainted task) and 'not perceived by the patient.
Ukrainian literary language is a complex communicative system, synchronous-diachronic section of which is informative regarding the state of development of national consciousness, intellectual level of society, and in a broader sense – concerning inscribing of the language in the history of national and world culture. This article is devoted to tracing the temporal and conceptual dynamics of the stylistic norm in the work of Ukrainian writers of the 20th century. The methodological basis for the study was made by such methods as comparison and analysis of literary works of Ukrainian writers of the XX century. As a result of the study, the authors concluded that the linguistic norm is a cognitive reference point for the scientific parameterization of the style norm of artistic discourse. The authors also emphasize that the dynamism of the stylistic artistic norm lies in the ability of the poetic language to respond to the development of artistic and linguistic consciousness, thinking of both the author and the reader in various manifestations.
Using multiple treebanks to improve parsing performance has shown positive results. However, to what extent similar, yet competing annotation decisions play in parser behavior is unclear. We investigate this within a multi-task learning (MTL) dependency parser setup on two parallel treebanks, UD and SUD, which, while possessing similar annotation schemes, differ in specific linguistic annotation preferences. We perform a set of experiments with different MTL architectural choices, comparing performance across various input embeddings. We find languages tend to pattern in loose typological associations, but generally the performance within an MTL setting is lower than single model baseline parsers for each annotation scheme. The main contributing factor seems to be the competing syntactic annotation information shared between treebanks in an MTL setting, which is shown in experiments against differently annotated treebanks. This suggests that the impact of how the signal is encoded for annotations and its influence on possible negative transfer is more important than that of the input embeddings in an MTL setting.
A growing body of research analyzing musical scores suggests mode’s relationship with other expressive cues has changed over time. However, to the best of our knowledge, the perceptual implications of these changes have not been formally assessed. Here, we explore how compositional choices of 17th- and 19th-century composers (J. S. Bach and F. Chopin, respectively) differentially affect emotional communication. This novel exploration builds on our team’s previous techniques using commonality analysis to decompose intercorrelated cues in unaltered excerpts of influential compositions. In doing so, we offer an important naturalistic complement to traditional experimental work—often involving tightly controlled stimuli constructed to avoid the intercorrelations inherent to naturalistic music. Our data indicate intriguing changes in cues’ effects between Bach and Chopin, consistent with score-based research suggesting mode’s “meaning” changed across historical eras. For example, mode’s unique effect accounts for the most variance in valence ratings of Chopin’s preludes, whereas its shared use with attack rate plays a more prominent role in Bach’s. We discuss the implications of these findings as part of our field’s ongoing effort to understand the complexity of musical communication—addressing issues only visible when moving beyond stimuli created for scientific, rather than artistic, goals.
State-of-the-art neural language models represented by Transformers are becoming increasingly complex and expensive for practical applications. Low-bit deep neural network quantization techniques provides a powerful solution to dramatically reduce their model size. Current low-bit quantization methods are based on uniform precision and fail to account for the varying performance sensitivity at different parts of the system to quantization errors. To this end, novel mixed precision DNN quantization methods are proposed in this paper. The optimal local precision settings are automatically learned using two techniques. The first is based on a quantization sensitivity metric in the form of Hessian trace weighted quantization perturbation. The second is based on mixed precision Transformer architecture search. Alternating direction methods of multipliers (ADMM) are used to efficiently train mixed precision quantized DNN systems. Experiments conducted on Penn Treebank (PTB) and a Switchboard corpus trained LF-MMI TDNN system suggest the proposed mixed precision Transformer quantization techniques achieved model size compression ratios of up to 16 times over the full precision baseline with no recognition performance degradation. When being used to compress a larger full precision Transformer LM with more layers, overall word error rate (WER) reductions up to 1.7% absolute (18% relative) were obtained.
The problems inherent in efforts to create large lexical databases by mapping machine-readable dictionaries onto each other are illustrated through an attempt to merge manually two short entries (whistle and whistler) from two monolingual English collegiate-style dictionaries. In view of the complexities revealed by this operation for the human mind, let alone the machine, we suggest that if lexical databases are to be created by automatic or semi-automatic means, it is essential to design the complete database first, following the demands of the language as identified by theoretical linguistic research, and only then to attempt to map the contents of any machine-readable dictionary into this ‘ideal’ database. We exemplify part of the suggested template lexical entry for verbs of sound, including whistle. 1
Postpartum Depression (PPD) is the most common non-obstetric complications associated with childbearing, but currently has poor diagnostic regimes. Sensory symptoms of PPD are understudied, particularly with regard to the sense of olfaction. The present study addresses this research gap by assessing differences in olfactory abilities between 39 depressed mothers, who were within the perinatal period (i.e., during pregnancy and up to 1-year post pregnancy) and assessed with Edinburgh Postnatal Depression Scale, and their case-matched healthy volunteers. The assessments include two olfactory testing sessions conducted 4-weeks apart, each comprising a standard odour detection threshold test (i.e., Snap & Sniff Olfactory Test System), and intensity and valence ratings for 3 "pleasant" and 3 "unpleasant" odorants. The results revealed no difference between patients (M = 5.6; SE = 0.3) and control group (M = 5.7; SE = 0.4) in terms of olfactory detection threshold. However, the patients group perceived the 3 "unpleasant" odours as significantly less pleasant (p < 0.05), and 2 odorants (1 "pleasant" and 1 "unpleasant") as less intense. Additionally, these results did not appear to be significantly interacted with the individual's perinatal stage. The present study is the first to evaluate associations between olfactory function and PPD. Findings from the study suggest that, while PPD has little effect on the early stages of olfactory processing, these conditions may have stronger influence on higher-order olfactory perception, including both hedonic and intensity perception. These novel findings add knowledge to sensory symptoms of PPD.
Emotion processing abnormalities and sleep pathology are central to the phenomenology of paediatric posttraumatic stress disorder, and sleep disturbance has been linked to the development, maintenance and severity of the disorder. Given emerging evidence indicating a role for sleep in emotional brain function, it has been proposed that dysfunctional processing of emotional experiences during sleep may play a significant role in affective disorders, including posttraumatic stress disorder. Here we sought to examine the relationship between sleep and emotion processing in typically developing youth, and youth with a diagnosis of posttraumatic stress disorder. We use high-density electroencephalogram to compare baseline sleep with sleep following performance on a task designed to assess both memory for and reactivity to negative and neutral imagery in 10 youths with posttraumatic stress disorder, and 10 age- and sex-matched non-traumatized typically developing youths. Subjective ratings of arousal to negative imagery (ΔArousal = post-sleep minus pre-sleep arousal ratings) remain unchanged in youth with posttraumatic stress disorder following sleep (mean increase 0.15, CI -0.28 to +0.58), but decreased in TD youth (mean decrease -1.0, 95% CI -1.44 to -0.58). ΔArousal, or affective habituation, was negatively correlated with global change in slow-wave activity power (ρ = -0.58, p =.008). When considered topographically, the correlation between Δslow-wave activity power and affective habituation was most significant in a frontal cluster of 27 electrodes (Spearman, ρ = -0.51, p =.021). Our results highlight the importance of slow-wave sleep for adaptive emotional processing in youth, and have implications for symptom persistence in paediatric posttraumatic stress disorder. Impairments in slow-wave activity may represent a modifiable risk factor in paediatric posttraumatic stress disorder.
Reproduction of knowledge, especially tacit knowledge can be expensive during a pandemic. One of the most common causes is the reduced information accessibility during the translation process. Having the ability to assess the linguistic complexity of any given contents could potentially improve knowledge reproduction. Authors conduct two cross-linguistic studies on the World Health Organization (WHO)'s emergency learning platform to assess the linguistic complexity of two online courses in 10 languages. Morpho-syntactically annotated treebanks, unannotated materials from Wikipedia and language-specific corpora are set as control groups. Preliminary findings reveal a clear reduced complexity of learning contents in the most candidate languages while retaining the maximum amount of information. Creating a baseline study on low-resourced languages on the learning genre could be potentially useful for measuring impact of normative products at country and local level.
BACKGROUND: Youth with anxiety disorders struggle with managing emotions relative to peers, but the neural basis of this difference has not been examined. METHODS: = 13.6; range = 8-17) with (n = 37) and without (n = 24) anxiety disorders completed a cognitive reappraisal task while undergoing functional magnetic resonance imaging. Emotional reactivity and regulation, functional activation, and beta-series connectivity were compared across groups. RESULTS: Groups did not differ on emotional reactivity or regulation. However, fronto-limbic activation after viewing aversive imagery with and without regulation, as well as affect ratings without regulation, were higher for anxious youth. Neither group demonstrated age-related changes in regulation, though anxious youth became less reactive with age. Stronger amygdala-ventromedial prefrontal cortex connectivity related to greater anxiety in control youth, but less anxiety in anxious youth. CONCLUSION: Anxious youth regulated when instructed, but regulation ability did not relate to age. Viewing aversive imagery related to heightened fronto-limbic activation even after reappraisal. Emotion dysregulation in youth anxiety disorders may stem from heightened emotionality and potent bottom-up neurobiological responses to aversive stimuli. Findings suggest the importance of treatments focused on both reducing initial emotional reactivity and bolstering regulatory capacity.
Recurrent neural networks are efficient ways of training language models, and various RNN networks have been proposed to improve performance. However, with the increase of network scales, the overfitting problem becomes more urgent. In this paper, we propose a framework-G2Basy-to speed up the training process and ease the overfitting problem. Instead of using predefined hyperparameters, we devise a gradient increasing and decreasing technique that changes the parameters training batch size and input dropout simultaneously by a user-defined step size. Together with a pretrained word embedding initialization procedure and the introduction of different optimizers at different learning rates, our framework speeds up the training process dramatically and improves performance compared with a benchmark model of the same scale. For the word embedding initialization, we propose the concept of "artificial features" to describe the characteristics of the obtained word embeddings. We experiment on two of the most often used corpora-the Penn Treebank and WikiText-2 datasets-and both outperform the benchmark results and show potential towards further improvement. Furthermore, our framework shows better results with the larger and more complicated WikiText-2 corpus than with the Penn Treebank. Compared with other state-of-the-art results, we achieve comparable results with network scales hundreds of times smaller and within fewer training epochs.
STUDY OBJECTIVES: Sleep plays a pivotal role in the off-line processing of emotional memory. However, much remains unknown for its immediate vs. long-term influences. We employed behavioral and electrophysiological measures to investigate the short- and long-term impacts of sleep vs. sleep deprivation on emotional memory. METHODS: Fifty-nine participants incidentally learned 60 negative and 60 neutral pictures in the evening and were randomly assigned to either sleep or sleep deprivation conditions. We measured memory recognition and subjective affective ratings in 12- and 60-h post-encoding tests, with EEGs in the delayed test. RESULTS: In a 12-h post-encoding test, compared to sleep deprivation, sleep equally preserved both negative and neutral memory, and their affective tones. In the 60-h post-encoding test, negative and neutral memories declined significantly in the sleep group, with attenuated emotional responses to negative memories over time. Furthermore, two groups showed spatial-temporally distinguishable ERPs at the delayed test: while both groups showed the old-new frontal negativity (300-500 ms, FN400), sleep-deprived participants additionally showed an old-new parietal, Late Positive Component effect (600-1000 ms, LPC). Multivariate whole-brain ERPs analyses further suggested that sleep prioritized neural representation of emotion over memory processing, while they were less distinguishable in the sleep deprivation group. CONCLUSIONS: These data suggested that sleep's impact on emotional memory and affective responses is time-dependent: sleep preserved memories and affective tones in the short term, while ameliorating affective tones in the long term. Univariate and multivariate EEG analyses revealed different neurocognitive processing of remote, emotional memories between sleep and sleep deprivation groups.
People use their previous experience to predict present affective events. Since we live in ever-changing environments, affective predictions must generalize from past contexts (from which they are implicitly learned) to new, potentially ambiguous contexts. This study investigated how past (un)certain relationships influence subjective experience following new ambiguous cues, and whether past relationships can be learned implicitly. Two S1-S2 paradigms were employed as learning and test phases in two experiments. S1s were colored circles, S2s negative or neutral affective pictures. Participants (N = 121, 116) were assigned to the certain (CG) or uncertain group (UG), and they were presented with 100% (CG) or 50% (UG) S1-S2 congruency during an uninstructed (Experiment 1) or implicit (Experiment 2) learning phase. During the test phase both groups were presented with a new 75% S1-S2 paradigm, and ambiguous (Experiment 1) or unambiguous (Experiment 2) S1s. Participants were asked to rate the expected valence of upcoming S2s (expectancy ratings), or their experienced valence and arousal (valence and arousal ratings). In Experiment 1 ambiguous cues elicited less negative expectancy ratings, and less unpleasant valence ratings, independently from prior experience. In Experiment 2, participants in the CG reported more negative expectancy ratings after the S1s previously paired with negative stimuli. Overall, we found that in the presence of ambiguous cues subjective affective experience is dampened, and we confirmed that people are able to infer probabilistic relationships from the environment (and to use them later) at an implicit level.
In this study, the affective explicit and implicit attitudes toward electric and gasoline cars are investigated. One hundred sixty-five participants (103 cisgender women, 62 cisgender men) completed an explicit and implicit affective rating task toward pictures of electric and gasoline cars, measurements of sustainability, future and past behaviors, and mindfulness. The results showed a positive emotional attitude for the electric cars compared with the gasoline cars only for the explicit rating but not for the implicit one. Furthermore, factors that correlated to the attitudes were investigated: explicit ratings in car owners correlated with age, degree, sustainability in general, and the expressed intention to purchase an electric car in the future. Implicit attitudes in car owners correlated with the overall score of mindfulness and the dimension of "non-reactivity." For the non-car owners, explicit attitudes correlated with the expressed intention to purchase an electric car in the future and the mindfulness dimension of "describing". In this group, the implicit attitude correlated negatively with the mindfulness intention of acting with awareness. This indicates that several different factors should be considered in the development of promotion campaigns for the advantage of sustainable mobility behavior.
OBJECTIVES: Age differences in affective experience across adulthood are widely documented. According to the circumplex model of affect consists of 2 aspects-valence (positive vs negative) and arousal (low activation vs high activation). Prior research on age differences has primarily focused on the valence aspect. However, little is known about age differences in daily affect of high and low arousal. METHOD: The present study examined age differences in daily dynamics (i.e., mean levels, variability, and inertia) of negative affect (NA) and positive affect (PA) of high and low arousal in a sample of 492 adults aged 21-91. Participants completed daily affect ratings for 21 consecutive days. RESULTS: Age was negatively and linearly related to mean levels of both high-arousal and low-arousal NA. Both high-arousal and low-arousal PA mean levels showed increases after middle age. Further, age was related to lower variability in both NA and PA regardless of arousal. Additionally, high-arousal NA inertia showed a linear decrease with age, whereas low-arousal PA inertia showed an inverted-U pattern with age. After controlling for mean levels of affect, the associations between age and affect variability remained significant, whereas the associations between age and affect inertia did not. DISCUSSION: The affective profile of older age is characterized by lower mean levels of NA, higher mean levels of PA, lower affect variability, and less persistence in high-arousal NA and low-arousal PA in daily life. Our results contribute to a nuanced understanding of which affective processes improve with age and which do not.
OBJECTIVE: To evaluate remote testing as a tool for measuring emotional responses to non-speech sounds. DESIGN: Participants self-reported their hearing status and rated valence and arousal in response to non-speech sounds on an Internet crowdsourcing platform. These ratings were compared to data obtained in a laboratory setting with participants who had confirmed normal or impaired hearing. STUDY SAMPLE: Adults with normal and impaired hearing. RESULTS: In both settings, participants with hearing loss rated pleasant sounds as less pleasant than did their peers with normal hearing. The difference in valence ratings between groups was generally smaller when measured in the remote setting than in the laboratory setting. This difference was the result of participants with normal hearing rating sounds as less extreme (less pleasant, less unpleasant) in the remote setting than did their peers in the laboratory setting, whereas no such difference was noted for participants with hearing loss. Ratings of arousal were similar from participants with normal and impaired hearing; the similarity persisted in both settings. CONCLUSIONS: In both test settings, participants with hearing loss rated pleasant sounds as less pleasant than did their normal hearing counterparts. Future work is warranted to explain the ratings of participants with normal hearing.
Background: Limited research has investigated whether replacing psychiatric diagnosis with psychological formulation-based approaches has implications for lay attitudes to mental health. The present study investigates experimentally whether presenting psychosis in terms of a schizophrenia diagnosis vs. formulation narrative affects stigma and treatment attitudes in the general public.Method: The study employed a between-groups experimental vignette design, with data collected online. 351 participants (64.1% female, aged 18–66, ) read a vignette about a person experiencing psychosis, defined with either a diagnosis of schizophrenia or a narrative-based formulation. Participants completed a battery of scales measuring their attitudes to the vignette character (social distance, attribution, recommended treatment options, mental help-seeking attitudes).Results: Desired social distance was significantly greater in participants exposed to the diagnostic label of schizophrenia. The schizophrenia label led participants to rate medical care as significantly more helpful relative to the formulation condition but did not affect ratings of specialist or community care or mental help-seeking attitudes.Conclusions: These findings suggest that a psychological formulation approach may slightly lessen stigma-related attitudes, relative to traditional diagnostic systems. Popularisation of formulation models need not compromise general orientations to help-seeking or perceived helpfulness of specialist care but may lead to less medicalised treatment preferences.
The PapyGreek Treebanks dataset contains documentary texts written in Postclassical Greek (ca. 300 BCE–700 CE), morphosyntactically annotated according to Dependency Grammar. The source of the texts is the Duke Databank of Documentary Papyri (DDbDP), which preserves the modern editorial treatment of the documents in TEI Epidoc XML encoding. Aiming to expose linguistic variation in the DDbDP, we have annotated two versions of a selection of documents: the plain transcription and an editorially corrected version. The dataset also comprises metadata about the documents’ dating and provenance, text type, and the persons involved. Furthermore, it facilitates linguistic research on these texts.
From a simplified analysis of adaptive methods, we derive AvaGrad, a new optimizer which outperforms SGD on vision tasks when its adaptability is properly tuned. We observe that the power of our method is partially explained by a decoupling of learning rate and adaptability, greatly simplifying hyperparameter search. In light of this observation, we demonstrate that, against conventional wisdom, Adam can also outperform SGD on vision tasks, as long as the coupling between its learning rate and adaptability is taken into account. In practice, AvaGrad matches the best results, as measured by generalization accuracy, delivered by any existing optimizer (SGD or adaptive) across image classification (CIFAR, ImageNet) and character-level language modelling (Penn Treebank) tasks. When training GANs, AvaGrad improves upon existing optimizers. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup>
Abstract The lexicon of emotion words is fundamental to interpersonal communication. To examine how emotion word acquisition interacts with societal context, the present study investigated emotion word development in three groups of child Korean users aged 4–13 years: those who use Korean primarily outside the home as a majority language (MajKCs) or inside the home as a minority language (MinKCs), and those who use Korean both inside and outside the home (KCs). These groups, along with a group of L1 Korean adults, rated the emotional valence of 61 Korean emotion words varying in frequency, valence, and age of acquisition. Results showed KCs, MajKCs, and MinKCs all converging toward adult-like valence ratings by ages 11–13 years; unlike KCs and MajKCs, however, MinKCs did not show age-graded development and continued to diverge from adults in emotion word knowledge by these later ages. These findings support the view that societal context plays a major role in emotion word development, offering one reason for the intergenerational communication difficulties reported by immigrant families.
Abstract In this paper we introduce an extended version of the Vedic Treebank ( vtb, Hellwig et al. 2020) which comes along with revisited and extended annotation guidelines. In order to assess the quality of our annotations as well as the usability and limits of the guidelines we performed an inter-annotator agreement test. The results show that agreement between annotators is hampered by various factors, most prominently by insufficient understanding of the content because of the cultural and temporal gap and incomplete knowledge of Vedic grammar. An in-depth discussion of disagreeing annotations demonstrates that the setup of the workflow, too, has a major influence on inter-annotator agreement. We suggest some measures that can help increase the transparency and annotation consistency according to current knowledge of the language when annotating Vedic Sanskrit, or ancient language varieties in general.
The paper investigates „Sprachliche Verrohung“ (linguistic neglection/brutalization), a term that has been recently and often used within the German mass media. It seems, however, that there is no common understanding of what „Sprachliche Verrohung“ is. To obtain a definition, a thinkaloud study was conducted: 40 participants judged relevant linguistic examples by verbalizing aloud their thoughts and ideas. The obtained think-aloud protocols are analysed and the following definition is derived: Expressions of „Sprachliche Verrohung“ are in conflict with the linguistic norm and speakers uses them despite their knowledge of this conflict.
Dictionary-based methods in sentiment analysis have received scholarly attention recently, the most comprehensive examples of which can be found in English.However, many other languages lack polarity dictionaries, or the existing ones are small in size as in the case of Senti-TurkNet, the first and only polarity dictionary in Turkish.Thus, this study aims to extend the content of SentiTurkNet by comparing the two available WordNets in Turkish, namely KeNet and TR-wordnet of BalkaNet.To this end, a current Turkish polarity dictionary has been created relying on 76,825 synsets matching KeNet, where each synset has been annotated with three polarity labels, which are positive, negative and neutral.Meanwhile, the comparison of KeNet and TR-wordnet of BalkaNet has revealed their weaknesses such as the repetition of the same senses, lack of necessary merges of the items belonging to the same synset and the presence of redundant narrower versions of synsets, which are discussed in light of their potential to the improvement of the current lexical databases of Turkish.
While educators may be well positioned to support unaccompanied immigrant youth, there is limited interdisciplinary research focused on understanding the complexity of youth’s experiences in US schools. The purpose of this qualitative, interview-based study was to better understand how youth’s transnational experiences pre-, during, and post-migration affected their school-based experiences, and to explore how schools supported them. Participants included ten unaccompanied immigrant youths from Central America and six key informants who worked with youth in a professional capacity. Findings indicate that youth experienced multiple challenges including stressful and traumatic events, barriers to mental health and legal services, and unfamiliar cultural and linguistic norms that sometimes were not recognized or understood by their teachers and schools. The youth also brought important resources, such as high expectations and aspirations and strong connections to family and community. School-based experiences that built from youth’s resources and motivations (e.g., through school-community partnerships and responsive classroom practices) had the potential to enhance belonging, community connections, and wellness. More interdisciplinary research is needed to develop and support school-based practices and partnerships in consultation with youth that build from knowledge of their particular resources and challenges.
While Out-of-distribution (OOD) detection has been well explored in computer\nvision, there have been relatively few prior attempts in OOD detection for NLP\nclassification. In this paper we argue that these prior attempts do not fully\naddress the OOD problem and may suffer from data leakage and poor calibration\nof the resulting models. We present PnPOOD, a data augmentation technique to\nperform OOD detection via out-of-domain sample generation using the recently\nproposed Plug and Play Language Model (Dathathri et al., 2020). Our method\ngenerates high quality discriminative samples close to the class boundaries,\nresulting in accurate OOD detection at test time. We demonstrate that our model\noutperforms prior models on OOD sample detection, and exhibits lower\ncalibration error on the 20 newsgroup text and Stanford Sentiment Treebank\ndataset (Lang, 1995; Socheret al., 2013). We further highlight an important\ndata leakage issue with datasets used in prior attempts at OOD detection, and\nshare results on a new dataset for OOD detection that does not suffer from the\nsame problem.\n
The complete semantic representation of a Tibetan sentence is mainly determined by the addition of a specific functional word. The choice of Tibetan functional words is mainly influenced (both explicitly and implicitly) by the sequence of Tibetan suffixes. In this article, we propose an RNN-based Tibetan radical suffix unit (TRSU) to consider this relationship. Specifically, for the Tibetan radical suffix unit-explicit (TRSU-E) method, the fixed suffix in Tibetan is used to determine the virtual functional words. For the Tibetan radical suffix unit-implicit (TRSU-I) method, the decision is assisted by adding a specific suffix. To test the method, we design a standard Tibetan corpus, which consists of different genres. Our experimental results show that the complexity of our method is reduced by up to 22.2% relative to the best baseline. Furthermore, with the hidden semantic information and implicit suffix, TRSU-I outperforms TRSU-E by reducing the perplexity (PPL) by 3%. Moreover, good results are achieved on the English Penn Treebank data set.
Facial expressions are a rich information source from which observers infer the emotional states of others. Despite much understanding about the brain regions that represent facial expressions, we do not yet know how representations of these facial movements transform into judgments of emotions in the brain. We addressed this question in 5 participants who judged the emotion of individual face movements called Action Units (AUs) while we concurrently measured brain activity using magnetoencephalography (MEG). Stimuli were animations of 5 facial movements--Outer Brower Raiser (AU2), Nose Wrinkler (AU9), Lip Corner Puller (AU12), Chin Raiser (AU17), Lip Stretcher (AU20), each at 4 levels of intensity (%25 - %100). We instructed participants to rate each animation according to either its perceived valence (‘negative’, ‘neutral’ or ‘positive’) or arousal (‘low,’ ‘neutral’ or ‘high’). Tasks alternated between blocks of 40 trials (5 AUs X 4 intensity levels X 2 repetitions) and participants completed 4,000 ~ 6,000 trials in total. We averaged all ratings of each AU and intensity level per task for each participant. We show that the arousal ratings increased along AU intensity levels while valence ratings are consistent for each AU (e.g., Nose Wrinkler (AU9) as negative and Lip Corner Puller (AU12) as positive). Then, we calculated Mutual Information (MI, permutation test) between MEG recording and task ratings. The results revealed the spatial and temporal distribution of brain activities related to the specific valence and arousal. We found that the valence and arousal evoked similar representational peaks ~270ms and ~750 ms in the temporal lobes while a special peak from parietal lobes at 387ms for valence task that differentiated between the two inferences. Our results show where (in temporal lobes and parietal lobes) and when (at ~270ms, 380ms and 750 ms post stimulus) the brain processes dynamic AUs as meaningful affective signals.