Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
16504 papers
This paper describes the system for our participation of team Wanghao-ftd-SJTU in the CoNLL 2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies. In this work, we design a system based on UDPipe 1 for universal dependency parsing, where transitionbased models are trained for different treebanks. Our system directly takes raw texts as input, performing several intermediate steps like tokenizing and tagging, and finally generates the corresponding dependency trees. For the special surprise languages for this task, we adopt a delexicalized strategy and predict based on transfer learning from other related languages. In the final evaluation of the shared task, our system achieves a result of 66.53% in macro-averaged LAS F1-score.
Recent work on the problem of latent tree learning has made it possible to train neural networks that learn to both parse a sentence and use the resulting parse to interpret the sentence, all without exposure to ground-truth parse trees at training time. Surprisingly, these models often perform better at sentence understanding tasks than models that use parse trees from conventional parsers. This paper aims to investigate what these latent tree learning models learn. We replicate two such models in a shared codebase and find that (i) only one of these models outperforms conventional tree-structured models on sentence classification, (ii) its parsing strategies are not especially consistent across random restarts, (iii) the parses it produces tend to be shallower than standard Penn Treebank (PTB) parses, and (iv) they do not resemble those of PTB or any other semantic or syntactic formalism that the authors are aware of.
Discourse parsing is an integral part of understanding information flow and argumentative structure in documents. Most previous research has focused on inducing and evaluating models from the English RST Discourse Treebank. However, discourse treebanks for other languages exist, including Spanish, German, Basque, Dutch and Brazilian Portuguese. The treebanks share the same underlying linguistic theory, but differ slightly in the way documents are annotated. In this paper, we present (a) a new discourse parser which is simpler, yet competitive (significantly better on 2/3 metrics) to state of the art for English, (b) a harmonization of discourse treebanks across languages, enabling us to present (c) what to the best of our knowledge are the first experiments on cross-lingual discourse parsing.
We propose to exploit the potential of multiple word clusterings in class-based recurrent neural network (RNN) language models for ensemble RNN language modeling. By varying the clustering criteria and the space of word embedding, different word clusterings are obtained to define different word/class factorizations. For each such word/class factorization, several base RNNLMs are learned, and the word prediction probabilities of the base RNNLMs are then combined to form an ensemble prediction. We use a greedy backward model selection procedure to select a subset of models and combine these models for word prediction. The proposed ensemble language modeling method has been evaluated on Penn Treebank test set as well as Wall Street Journal (WSJ) Eval 92 and 93 test sets, where it improved test set perplexity and word error rate over the state-of-the-art single RNNLMs as well as multiple RNNLMs produced by varying RNN learning conditions.
This paper attempts to submit an evaluation plan based on word naming tes perspectives. The uniqueness of the evaluation of this perspective is the use of psycholinguistic norms to identify inhibiting factors as well as supporting kanji recognition by learners. In addition, a detailed analysis of the kanji constituents is to be tested. The psycholinguistic norms and characteristics of the kanji constituents are exemplified in the design of the evaluation in this paper, including: frequency effects, cumulative frequency effects, family size effects, word complexity, and so on. While the kanji constituent seen from the visual form of kanji graphics on the phonetic and semantic radicals; a nd consistency of compound kanji pronunciation. This evaluation is based on a study of how information processing takes place in the mind of the learners. In the field of visual recognition of words, the task commonly used to test cognitive learning abilities is the word naming test. In this test the learners are asked to pronounce the visible kanji on the computer monitor quickly and accurately within the specified response time limit. Learners' answers will be recorded and then analyzed offline. Advantage of this evaluation is teacher or lecturer can identify factors causing learners difficulties in recognizing kanji. Keywords: word naming tes, ability to read kanji, evaluation
A method is presented which applies Long Short-Term Memory Recurrent Neural Networks on real market-research voice recordings in order to automatically predict emotional arousal from speech. While most previous work has dealt with evaluations of algorithms within the same speech corpus, the novelty of this paper lies in an extensive evaluation across corpora and languages. The approach is evaluated on seven large data sets collected in real tests of TV commercials and new product concepts across four languages. We observe excellent performance within and between the different corpora when compared against the gold standard of arousal ratings by human annotators. Even in the cross-language validation the models show good performance which almost reaches human rater agreement.
Universal Dependency (UD) annotations, despite their usefulness for cross-lingual tasks and semantic applications, are not optimised for statistical parsing.In the paper, we ask what exactly causes the decrease in parsing accuracy when training a parser on UD-style annotations and whether the effect is similarly strong for all languages.We conduct a series of experiments where we systematically modify individual annotation decisions taken in the UD scheme and show that this results in an increased accuracy for most, but not for all languages.We show that the encoding in the UD scheme, in particular the decision to encode content words as heads, causes an increase in dependency length for nearly all treebanks and an increase in arc direction entropy for many languages, and evaluate the effect this has on parsing accuracy.
Higher emotional clarity, the extent to which people unambiguously identify, label, and describe their own emotions, is related to a host of positive intrapersonal factors but its relation to interpersonal factors is unexplored. We hypothesized that emotional clarity would be related to cognitive empathy (i.e., perceiving others' emotions) and to accurately understanding others' negative affect (NA), but not positive affect (PA), in the context of a stressful situation. After completing self-reports of trait emotional clarity and cognitive and affective empathy (i.e., one's emotional reaction to others), participants (N = 94 undergraduate students; i.e., perceivers) viewed a series of video clips of adults (i.e., targets) completing a stressful laboratory task in a previous research study. Before and after the stress task, targets reported their state NA and PA. While viewing the recordings, perceivers rated how they thought the targets were feeling at the corresponding time points. Correspondence between perceivers' and targets' affect ratings were used as indices of the outcome variable, performance-based cognitive empathy. As expected, self-reported emotional clarity was related to the self-reported cognitive, but not affective, empathy. Moreover, perceivers' emotional clarity was related to higher cognitive empathy for NA not PA after the stressful task. Our findings provide preliminary support for the importance of emotional clarity in the ability to accurately understand others' affective experiences, which has important interpersonal implications. (PsycINFO Database Record
This paper investigates interactions in parser performance for the two official standards for written Norwegian: Bokmål and Nynorsk. We demonstrate that while applying models across standards yields poor performance, combining the training data for both standards yields better results than previously achieved for each of them in isolation. This has immediate practical value for processing Norwegian, as it means that a single parsing pipeline is sufficient to cover both varieties, with no loss in accuracy. Based on the Norwegian Universal Dependencies treebank we present results for multiple taggers and parsers, experimenting with different ways of varying the training data given to the learners, including the use of machine translation.
This paper aims at questioning two diasporic communities settled in Noumea, New Caledonia's capital city. The first consists of those with internal (domestic) mobility (from Lifou and Ouvea, Loyalty Islands), while the second are those from Vanuatu, the product of external (inter-regional) migration. From these case studies, we examine how unevenly they are managed bytthe weight of linguistic norms on everyday language practices of diasporic urban speakers of Drehu, Iaai and Bislama languages. This contribution also investigates the strategies used by diasporic speakers as well as language institutions, like the Kanak Languages Academy, to maintain their languages and preserve their heritage.
Generative neural models have recently achieved state-of-the-art results for constituency parsing. However, without a feasible search procedure, their use has so far been limited to reranking the output of external parsers in which decoding is more tractable. We describe an alternative to the conventional action-level beam search used for discriminative neural models that enables us to decode directly in these generative models. We then show that by improving our basic candidate selection strategy and using a coarse pruning function, we can improve accuracy while exploring significantly less of the search space. Applied to the model of Choe and Charniak (2016), our inference procedure obtains 92.56 F1 on section 23 of the Penn Treebank, surpassing prior state-of-the-art results for single-model systems.
Discourse-annotated corpora are an important resource for the community, but they are often annotated according to different frameworks. This makes comparison of the annotations difficult, thereby also preventing researchers from searching the corpora in a unified way, or using all annotated data jointly to train computational systems. Several theoretical proposals have recently been made for mapping the relational labels of different frameworks to each other, but these proposals have so far not been validated against existing annotations. The two largest discourse relation annotated resources, the Penn Discourse Treebank and the Rhetorical Structure Theory Discourse Treebank, have however been annotated on the same text, allowing for a direct comparison of the annotation layers. We propose a method for automatically aligning the discourse segments, and then evaluate existing mapping proposals by comparing the empirically observed against the proposed mappings. Our analysis highlights the influence of segmentation on subsequent discourse relation labeling, and shows that while agreement between frameworks is reasonable for explicit relations, agreement on implicit relations is low. We identify several sources of systematic discrepancies between the two annotation schemes and discuss consequences of these discrepancies for future annotation and for the training of automatic discourse relation labellers.
We present a simple LSTM-based transition-based dependency parser. Our model is composed of a single LSTM hidden layer replacing the hidden layer in the usual feed-forward network architecture. We also propose a new initialization method that uses the pre-trained weights from a feed-forward neural network to initialize our LSTM-based model. We also show that using dropout on the input layer has a positive effect on performance. Our final parser achieves a 93.06% unlabeled and 91.01% labeled attachment score on the Penn Treebank. We additionally replace LSTMs with GRUs and Elman units in our model and explore the effectiveness of our initialization method on individual gates constituting all three types of RNN units.
We explore the concept of hybrid grammars, which formalize and generalize a range of existing frameworks for dealing with discontinuous syntactic structures. Covered are both discontinuous phrase structures and non-projective dependency structures. Technically, hybrid grammars are related to synchronous grammars, where one grammar component generates linear structures and another generates hierarchical structures. By coupling lexical elements of both components together, discontinuous structures result. Several types of hybrid grammars are characterized. We also discuss grammar induction from treebanks. The main advantage over existing frameworks is the ability of hybrid grammars to separate discontinuity of the desired structures from time complexity of parsing. This permits exploration of a large variety of parsing algorithms for discontinuous structures, with different properties. This is confirmed by the reported experimental results, which show a wide variety of running time, accuracy, and frequency of parse failures.
We tackle the prediction of instructor intervention in student posts from discussion forums in Massive Open Online Courses (MOOCs). Our key finding is that using automatically obtained discourse relations improves the prediction of when instructors intervene in student discussions, when compared with a state-of-the-art, feature-rich baseline. Our supervised classifier makes use of an automatic discourse parser which outputs Penn Discourse Treebank (PDTB) tags that represent in-post discourse features. We show PDTB relation-based features increase the robustness of the classifier and complement baseline features in recalling more diverse instructor intervention patterns. In comprehensive experiments over 14 MOOC offerings from several disciplines, the PDTB discourse features improve performance on average. The resultant models are less dependent on domain-specific vocabulary, allowing them to better generalize to new courses.
Representing multiple compositions of human language has been a difficult task due to the complex hierarchical and compositional nature of language. Hierarchical structures are one of the architectures which can be used to capture such compositionalities. In this paper, we introduce temporal hierarchies to the Neural Language Model (NLM) with the help of a Deep Gated Recurrent Neural Network with adaptive timescales to help represent multiple compositions of language. We demonstrate that by representing multiple compositions of language in a deep recurrent neural network architecture, we can improve the performance of Language Models without complex hierarchical architectures. We report the performance of the proposed model using the popular Penn Treebank (PTB) dataset. The results show that by using the multiple timescale concept in an NLM, we can achieve better perplexities compared to the existing baselines.
Unsupervised dependency parsing aims to learn a dependency parser from unannotated sentences. Existing work focuses on either learning generative models using the expectation-maximization algorithm and its variants, or learning discriminative models using the discriminative clustering algorithm. In this paper, we propose a new learning strategy that learns a generative model and a discriminative model jointly based on the dual decomposition method. Our method is simple and general, yet effective to capture the advantages of both models and improve their learning results. We tested our method on the UD treebank and achieved a state-ofthe-art performance on thirty languages.
We present the TurkuNLP entry in the CoNLL 2017 Shared Task on Multilingual Parsing from Raw Text to Universal Dependencies. The system is based on the UDPipe parser with our focus being in exploring various techniques to pre-train the word embeddings used by the parser in order to improve its performance especially on languages with small training sets. The system ranked 11th among the 33 participants overall, being 8th on the small treebanks, 10th on the large treebanks, 12th on the parallel test sets, and 26th on the surprise languages.
This article is about concept of electronic service that detects violations of normality in the creating and using of terminology units and provides recommendations for term using. This concept is adapted for specialists in electric power engineering and can be used for websites about this industry. Using these websites ensure compliance with linguistic norms in professional communication. Scheme of functioning of terminological online assistant for electric power engineering specialists is developed. The search and terms adding algorithm of terminological online assistant for electric power engineering specialists is described in this article. Entering a term into the search box, checking normality of term using database system, proposing list of recommended term equivalent, adding term to database system and giving results are the main components of developed algorithm.
It is widely believed that different parts of a classical Chinese poem vary in syntactic properties. The middle part is usually parallel, i.e. the two lines in a couplet have similar sentence structure and part of speech; in contrast, the beginning and final parts tend to be non-parallel. Imagistic language, dominated by noun phrases evoking images, is concentrated in the middle; propositional language, with more complex grammatical structures, is more often found at the end. We present the first quantitative analysis on these linguistic phenomena—syntactic parallelism, imagistic language, and propositional language—on a treebank of selected poems from the Complete Tang Poems. Written during the Tang Dynasty between the 7th and 9th centuries CE, these poems are often considered the pinnacle of classical Chinese poetry. Our analysis affirms the traditional observation that the final couplet is rarely parallel; the middle couplets are more frequently parallel, especially at the phrase rather than the word level. Further, the final couplet more often takes a non-declarative mood, uses function words, and adopts propositional language. In contrast, the beginning and middle couplets employ more content words and tend toward imagistic language.
In this study, we presented pictorial representations of happy, neutral, and fearful expressions projected in the eye regions to determine whether the eye region alone is sufficient to produce a context effect. Participants were asked to judge the valence of surprised faces that had been preceded by a picture of an eye region. Behavioral results showed that affective ratings of surprised faces were context dependent. Prime-related ERPs with presentation of happy eyes elicited a larger P1 than those for neutral and fearful eyes, likely due to the recognition advantage provided by a happy expression. Target-related ERPs showed that surprised faces in the context of fearful and happy eyes elicited dramatically larger C1 than those in the neutral context, which reflected the modulation by predictions during the earliest stages of face processing. There were larger N170 with neutral and fearful eye contexts compared to the happy context, suggesting faces were being integrated with contextual threat information. The P3 component exhibited enhanced brain activity in response to faces preceded by happy and fearful eyes compared with neutral eyes, indicating motivated attention processing may be involved at this stage. Altogether, these results indicate for the first time that the influence of isolated eye regions on the perception of surprised faces involves preferential processing at the early stages and elaborate processing at the late stages. Moreover, higher cognitive processes such as predictions and attention can modulate face processing from the earliest stages in a top-down manner.
OBJECTIVE: New MRI sequences based on rapid radial acquisition have reduced gradient noise. The purpose of this study was to compare Silent T1-weighted and unenhanced MR angiography (MRA) against conventional sequences in a clinical population. MATERIALS AND METHODS: The study cohort consisted of 40 patients with suspected brain metastases (median age, 60 years; range, 23-91 years) who underwent T1-weighted contrast-enhanced MRI and 51 patients with suspected vascular lesions or cerebral ischemia (median age, 60 years; range, 16-94 years) who underwent unenhanced intracranial MRA. Three neuroradiologists reviewed the images blindly and rated several measures of image quality on a 5-point Likert scale. Reviewers recorded the number of enhancing lesions and whether Silent images were better than, worse than, or equivalent to conventional images. RESULTS: For T1-weighted MR images, ratings were slightly lower for Silent versus conventional images, except for diagnostic confidence. Although more lesions were detected on conventional images, this difference was not statistically significant; agreement was seen in 88% of cases. In 48% of cases, T1-weighted scans were deemed equivalent, but when a preference existed, it was usually for conventional images (38% vs 14%). Conventional MRA images were rated higher on all image quality metrics and were strongly preferred (reviewers preferred conventional images in 69% of cases, rated the images as equivalent in 27% of cases, and preferred Silent images in 4% of cases). In some cases, artifacts on Silent images caused reduced vessel caliber, vessel irregularities, and even absent vessels. CONCLUSION: Although conventional T1-weighted images were preferred overall, most Silent T1-weighted images were rated as equivalent to or better than conventional images and represent a potential alternative for imaging of noise-averse patients. Silent MRA scored significantly worse and could not be recommended at this time, suggesting that it requires additional refinement before routine clinical use.
This paper formalizes a sound extension of dynamic oracles to global training, in the frame of transition-based dependency parsers. By dispensing with the precomputation of references, this extension widens the training strategies that can be entertained for such parsers; we show this by revisiting two standard training procedures, early-update and max-violation, to correct some of their search space sampling biases. Experimentally, on the SPMRL treebanks, this improvement increases the similarity between the train and test distributions and yields performance improvements up to 0.7 UAS, without any computation overhead.
The availability of the Rhetorical Structure Theory (RST) Discourse Treebank has spurred substantial research into discourse analysis of written texts; however, limited research has been conducted to date on RST annotation and parsing of spoken language, in particular, nonnative spontaneous speech. Considering that the measurement of discourse coherence is typically a key metric in human scoring rubrics for assessments of spoken language, we initiated a research effort to obtain RST annotations of a large number of non-native spoken responses from a standardized assessment of academic English proficiency. The resulting inter-annotator agreements on the three different levels of Span, Nuclearity, and Relation are 0.848, 0.766, and 0.653, respectively. Furthermore, a set of features was explored to evaluate the discourse structure of non-native spontaneous speech based on these annotations; the highest performing feature showed a correlation of 0.612 with scores of discourse coherence provided by expert human raters.
The main objective of this research was to study the comprehension level of heritage speakers of Turkish with regard to Turkish proverbs. The familiarity factor in relation to its role in comprehending proverbs is also examined. Familiarity judgments of proverbs were made by heritage speakers of Turkish to determine whether the familiarity ratings of the heritage speakers of Turkish could be associated with their understanding of proverbs. The results of the proverb comprehension tests indicate that the difference between the bilingual heritage speakers of Turkish and baseline monolingual speakers in their comprehension of proverbs was significant. Yet, when the arithmetic mean rank is taken into consideration, both heritage speakers of Turkish and monolinguals display high level performance in their comprehension of proverbs. However, there appears to be a noteworthy difference between the frequency level of heritage speakers’ and monolinguals’ in their encounter with proverbs. Monolinguals outperformed bilinguals in comprehension of proverbs and familiarity rating. Results also showed that familiarity had a nonsignificant correlation with participants’ performance on proverb comprehension.
OBJECTIVE: Previous studies have reported that brain-injured patients frequently suffer from cognitive impairments such as attention and concentration deficits. Numerous rehabilitation clinics offer animal-assisted therapy (AAT) to address these difficulties. The authors' aim was to investigate the immediate effects of AAT on the concentration and attention span of brain-injured patients. METHOD: Nineteen patients with acquired brain injury were included in a randomized, controlled, within-subject trial. The patients alternately received 12 standard therapy sessions (speech therapy, physiotherapy, occupational therapy) and 12 paralleled AAT sessions with comparable content. A total of 429 therapy sessions was analyzed consisting of 214 AAT and 215 control sessions. Attention span and instances of distraction were assessed via video coding in Noldus Observer. The Mehrdimensionaler Befindlichkeitsbogen ([Multidimensional Affect Rating Scale] MDBF questionnaire; Steyer, Schwenkmezger, Notz, & Eid, 1997) was used to measure the patient's self-rated alertness. Concentration was assessed through Visual Analogue Scale (VAS) via self-assessment and therapist's ratings. RESULTS: The patients' attention span did not differ whether an animal was present or not. However, patients displayed more instances of distraction during AAT. Moreover, patients rated themselves more concentrated and alert during AAT sessions. Further, therapists' evaluation of patients' concentration indicated that patients were more concentrated in AAT compared with the control condition. CONCLUSIONS: Although the patients displayed more instances of distraction while in the presence of an animal, it did not have a negative impact on their attention span. In addition, patients reported to be more alert and concentrated when an animal was present. Future studies should examine other attentional processes such as divided attention and include neurobiological correlates of attention. (PsycINFO Database Record
Two experiments examined if exposure to emotionally valent image-based secondary tasks introduced at different points of a free recall working memory (WM) task impair memory performance. Images from the International Affective Picture System (IAPS) varied in the degree of negative or positive valance (mild, moderate, strong) and were positioned at low, moderate and high WM load points with participants rating them based upon perceived valence. As predicted, and based on previous research and theory, the higher the degree of negative (Experiment 1) and positive (Experiment 2) valence and the higher the WM load when a secondary task was introduced, the greater the impairment to recall. Secondary task images with strong negative valance were more disruptive than negative images with lower valence at moderate and high WM load task points involving encoding and/or rehearsal of primary task words (Experiment 1). This was not the case for secondary tasks involving positive images (Experiment 2), although participant valence ratings for positive IAPS images classified as moderate and strong were in fact very similar. Implications are discussed in relation to research and theory on task interruption and attentional narrowing and literature concerning the effects of emotive stimuli on cognition.
Speech community concept can be seen/shown through three approaches; Firstly, on the basis of their common language forms, secondly by rules that regulate those common language forms, and thirdly from the perspective of the common cultural concepts. Speech community is possibly created if certain group of individuals, for the reason of common territory, professions, hobbies have exactly the same language and possess the common judgement on any linguistic norms. The same case can also be projected to any speech communities in certain social, household, governmental, or religious domains to mention a few possibilitites. Keywords: linguis politnes, sociolinguistic, speech commuity.
Describing implicit phenomena in discourse is known to be a problematic task, from both theoretical and empirical perspectives. The present article contributes to this topic by a novel comparative analysis of two prominent annotation approaches to discourse relations (coherence relations) that were carried out on the same texts. We compare the annotation of implicit relations in the Penn Discourse Treebank 2.0, i.e. discourse relations not signaled by an explicit discourse connective, to the recently released analysis of signals of rhetorical relations in the RST Signalling Corpus (RST-SC). The intersection of corresponding pairs of relations is rather a small one, but it shows a clear tendency: unlike the overall signal distribution in the RST-SC, more than half of the signals in the studied intersection are of semantic type, formed mostly by loosely defined lexical chains. Our data transformation allows for a simultaneous depiction and detailed study of the two resources.
Most fetal brain MRI reconstruction algorithms rely only on brain tissue-relevant voxels of low-resolution (LR) images to enhance the quality of inter-slice motion correction and image reconstruction. Consequently the fetal brain needs to be localized and extracted as a first step, which is usually a laborious and time consuming manual or semi-automatic task. We have proposed in this work to use age-matched template images as prior knowledge to automatize brain localization and extraction. This has been achieved through a novel automatic brain localization and extraction method based on robust template-to-slice block matching and deformable slice-to-template registration. Our template-based approach has also enabled the reconstruction of fetal brain images in standard radiological anatomical planes in a common coordinate space. We have integrated this approach into our new reconstruction pipeline that involves intensity normalization, inter-slice motion correction, and super-resolution (SR) reconstruction. To this end we have adopted a novel approach based on projection of every slice of the LR brain masks into the template space using a fusion strategy. This has enabled the refinement of brain masks in the LR images at each motion correction iteration. The overall brain localization and extraction algorithm has shown to produce brain masks that are very close to manually drawn brain masks, showing an average Dice overlap measure of 94.5%. We have also demonstrated that adopting a slice-to-template registration and propagation of the brain mask slice-by-slice leads to a significant improvement in brain extraction performance compared to global rigid brain extraction and consequently in the quality of the final reconstructed images. Ratings performed by two expert observers show that the proposed pipeline can achieve similar reconstruction quality to reference reconstruction based on manual slice-by-slice brain extraction. The proposed brain mask refinement and reconstruction method has shown to provide promising results in automatic fetal brain MRI segmentation and volumetry in 26 fetuses with gestational age range of 23 to 38 weeks.
Connections play a crucial role in neural network (NN) learning because they determine how information flows in NNs. Suitable connection mechanisms may extensively enlarge the learning capability and reduce the negative effect of gradient problems. In this paper, a new delay connection is proposed for Long Short-Term Memory (LSTM) unit to develop a more sophisticated recurrent unit, called Delay Connected LSTM (DCLSTM). The proposed delay connection brings two main merits to DCLSTM with introducing no extra parameters. First, it allows the output of the DCLSTM unit to maintain LSTM, which is absent in the LSTM unit. Second, the proposed delay connection helps to bridge the error signals to previous time steps and allows it to be back-propagated across several layers without vanishing too quickly. To evaluate the performance of the proposed delay connections, the DCLSTM model with and without peephole connections was compared with four state-of-the-art recurrent model on two sequence classification tasks. DCLSTM model outperformed the other models with higher accuracy and F1[Formula: see text]score. Furthermore, the networks with multiple stacked DCLSTM layers and the standard LSTM layer were evaluated on Penn Treebank (PTB) language modeling. The DCLSTM model achieved lower perplexity (PPL)/bit-per-character (BPC) than the standard LSTM model. The experiments demonstrate that the learning of the DCLSTM models is more stable and efficient.
A better understanding of factors that differentiate those who only experience suicidal ideation from those who engage in self-directed violence (SDV) is critical for suicide prevention efforts (Klonsky & May, 2014; May & Klonsky, 2016). To identify who is at greatest risk for death by suicide, it is imperative that new innovative assessment tools be created to facilitate behavioral measurement of key constructs associated with increased risk for SDV. The aim of the current study was to develop and validate a set of suicide-specific images, called the Self-Directed Violence Picture System (SDVPS), to help meet this need. A sample of 119 U.S. military veterans provided valence, arousal, and dominance ratings on the SDVPS. These ratings were compared to International Affective Picture System (IAPS) negative, neutral, and positive images. SDVPS images were rated with significantly greater negative valence and elicited decreased feelings of being in control than did IAPS positive (p <.001, p <.001), IAPS negative (p =.03, p =.001), and IAPS neutral (p <.001, p <.001) images. SDVPS images were also rated with significantly greater arousal than were IAPS neutral images (p <.001). Initial validation data support that the SDVPS images functioned as intended. Although continued validation of the SDVPS in other populations is necessary, the SDVPS may become a new tool by which researchers can begin to systematically and reliably examine reactions to suicide-related content using behavioral and/or experimental paradigms. (PsycINFO Database Record
This paper reports on a suite of experiments that evaluates how the linguistic granularity of part-of-speech tagsets impacts the performance of tagging and syntactic dependency parsing. Our results show that parsing accuracy can be significantly improved by introducing more finegrained morphological information in the tagset, even if tagger accuracy is compromised. Our taggers and parsers are trained and tested using the annotations of the Norwegian Dependency Treebank.
This paper describes the Amobee sentiment analysis system, adapted to compete in SemEval 2017 task 4. The system consists of two parts: a supervised training of RNN models based on a Twitter sentiment treebank, and the use of feedforward NN, Naive Bayes and logistic regression classifiers to produce predictions for the different sub-tasks. The algorithm reached the 3rd place on the 5-label classification task (sub-task C).
142:337-350), this study examined whether these abnormalities also characterize individuals at clinical high risk for MDD. We systematically explored the impact of family risk status and personal history of depression and anxiety on three distinct stages of emotional processing comprising the late positive potential (LPP). ERPs (72 channels) were recorded from 74 high and 53 low risk individuals (age 13-59 years, 58 male) during a visual half-field paradigm using highly-controlled pictures of cosmetic surgery patients showing disordered (negative) or healed (neutral) facial areas before or after treatment. Reference-free current source density (CSD) transformations of ERP waveforms were quantified by temporal principal components analysis (tPCA). Component scores of prominent CSD-tPCA factors sensitive to emotional content were analyzed via permutation tests and repeated measures ANOVA for mixed factorial designs with unstructured covariance matrix, including gender, age and clinical covariates. Factor-based distributed inverse solutions provided descriptive estimates of emotional brain activations at group level corresponding to hierarchical activations along ventral visual processing stream. Risk status affected emotional responsivity (increased positivity to negative-than-neutral stimuli) overlapping early N2 sink (peak latency 212 ms), P3 source (385 ms), and a late centroparietal source (630 ms). High risk individuals had reduced right-greater-than-left emotional lateralization involving occipitotemporal cortex (N2 sink) and bilaterally reduced emotional effects involving posterior cingulate (P3 source) and inferior temporal cortex (630 ms) when compared to those at low risk. While the early emotional effects were enhanced for left hemifield (right hemisphere) presentations, hemifield modulations did not differ between risk groups, suggesting top-down rather than bottom-up effects of risk. Groups did not differ in their stimulus valence or arousal ratings. Similar effects were seen for individuals with a lifetime history of depression or anxiety disorder in comparison to those without. However, there was no evidence that risk status and history of MDD or anxiety disorder interacted in their impact on emotional responsivity, suggesting largely independent attenuation of attentional resource allocation to enhance perceptual processing of motivationally salient stimuli. These findings further suggest that a deficit in motivated attention preceding conscious awareness may be a marker of risk for depression.
Humans are sensitive to gaze direction from early life, and gaze has social and affective values. Borderline personality disorder (BPD) is a clinical condition characterized by emotional dysregulation and enhanced sensitivity to affective and social cues. In this study we wanted to investigate the temporal-spatial dynamics of spontaneous gaze processing in BPD. We used a 2-back-working-memory task, in which neutral faces with direct and averted gaze were presented. Gaze was used as an emotional modulator of event-related-potentials to faces. High density EEG data were acquired in 19 females with BPD and 19 healthy women, and analyzed with a spatio-temporal microstates analysis approach. Independently of gaze direction, BPD patients showed altered N170 and P200 topographies for neutral faces. Source localization revealed that the anterior cingulate and other prefrontal regions were abnormally activated during the N170 component related to face encoding, while middle temporal deactivations were observed during the P200 component. Post-task affective ratings showed that BPD patients had difficulty to disambiguate neutral gaze. This study provides first evidence for an early neural bias toward neutral faces in BPD independent of gaze direction and also suggests the importance of considering basic aspects of social cognition in identifying biological risk factors of BPD.
The PARSEME shared task aims at identifying verbal MWEs in running texts. Verbal MWEs include idioms (let the cat out of the bag), light verb constructions (make a decision), verb-particle constructions (give up), and inherently reflexive verbs (se suicider 'to suicide' in French). VMWEs were annotated according to the universal guidelines in 18 languages. The corpora are provided in the parsemetsv format, inspired by the CONLL-U format. For most languages, paired files in the CONLL-U format - not necessarily using UD tagsets - containing parts of speech, lemmas, morphological features and/or syntactic dependencies are also provided. Depending on the language, the information comes from treebanks (e.g., Universal Dependencies) or from automatic parsers trained on treebanks (e.g., UDPipe). This item contains training and test data, tools and the universal guidelines file.
Singlish can be interesting to the ACL community both linguistically as a major creole based on English, and computationally for information extraction and sentiment analysis of regional social media. We investigate dependency parsing of Singlish by constructing a dependency treebank under the Universal Dependencies scheme, and then training a neural network model by integrating English syntactic knowledge into a state-of-the-art parser trained on the Singlish treebank. Results show that English knowledge can lead to 25% relative error reduction, resulting in a parser of 84.47% accuracies. To the best of our knowledge, we are the first to use neural stacking to improve cross-lingual dependency parsing on low-resource languages. We make both our annotation and parser available for further research.
Discourse relations can either be explicitly marked by discourse connectives (DCs), such as therefore and but, or implicitly conveyed in natural language utterances. How speakers choose between the two options is a question that is not well understood. In this study, we propose a psycholinguistic model that predicts whether or not speakers will produce an explicit marker given the discourse relation they wish to express. Our model is based on two information-theoretic frameworks: (1) the Rational Speech Acts model, which models the pragmatic interaction between language production and interpretation by Bayesian inference, and (2) the Uniform Information Density theory, which advocates that speakers adjust linguistic redundancy to maintain a uniform rate of information transmission. Specifically, our model quantifies the utility of using or omitting a DC based on the expected surprisal of comprehension, cost of production, and availability of other signals in the rest of the utterance. Experiments based on the Penn Discourse Treebank show that our approach outperforms the state-of-the-art performance at predicting the presence of DCs (Patterson and Kehler, 2013), in addition to giving an explanatory account of the speaker’s choice.
The negativity bias has been shown in many fields, including in face processing. We assume that this bias stems from the potential threat inlayed in the stimuli (e.g., negative moral behaviors) in previous studies. In the present study, we conducted one behavioral and one event-related potentials (ERPs) experiments to test whether the positivity bias rather than negativity bias will arise when participants process information whose negative aspect involves no threat, i.e., the ability information. In both experiments, participants first completed a valence rating (negative-to-positive) of neutral facial expressions. Further, in the learning period, participants associated the neutral faces with high-ability, low-ability, or control sentences. Finally, participants rated these facial expressions again. Results of the behavioral experiment showed that compared with pre-learning, the expressions of the faces associated with high ability sentences were classified as more positive in the post-learning expression rating task, and the faces associated with low ability sentences were evaluated as more negative. Meanwhile, the change in the high-ability group was greater than that of the low-ability group. The ERP data showed that the faces associated with high-ability sentences elicited a larger early posterior negativity, an ERP component considered to reflect early sensory processing of the emotional stimuli, than the faces associated with control sentences. However, no such effect was found in faces associated with low-ability sentences. To conclude, high ability sentences exerted stronger influence on expression perception than did low ability ones. Thus, we found a positivity bias in this ability-related facial perceptual task. Our findings demonstrate an effect of valenced ability information on face perception, thereby adding to the evidence on the opinion that person-related knowledge can influence face processing. What's more, the positivity bias in non-threatening surroundings increases scope for studies on processing bias.
This paper is concerned with whether deep syntactic information can help surface parsing, with a particular focus on empty categories. We design new algorithms to produce dependency trees in which empty elements are allowed, and evaluate the impact of information about empty category on parsing overt elements. Such information is helpful to reduce the approximation error in a structured parsing model, but increases the search space for inference and accordingly the estimation error. To deal with structure-based overfitting, we propose to integrate disambiguation models with and without empty elements, and perform structure regularization via joint decoding. Experiments on English and Chinese TreeBanks with different parsing models indicate that incorporating empty elements consistently improves surface parsing.
In this paper, an effective machine translation system from Thai to Khmer language on a website is proposed. To create a web application for a high performance Thai-Khmer machine translation (ThKh-MT), the principles and methods of translation involve with lexical base. Word reordering is applied by considering the previous word, the next word and subject-verb agreement. The word adjustment is also required to attain acceptable outputs. Additional steps related to structure patterns are added in a combination with the classical methods to deal with translation issues. PHP is implemented to build the application with MySQL as a tool to create lexical databases. For testing, 5,100 phrases and sentences are selected to evaluate the system. The result shows 89.25 percent of accuracy and 0.84 for F-Measure which infers to a higher efficiency than that of Google and other systems.
Cultural differences may influence interactions between humans with different social norms and cultural traits, incurring different emotional and behavioral responses. The same applies to human-robot interaction (HRI). We believe that controlling robot emotions based on the cultural context can help robots adapt to humans from culturally diverse backgrounds. Such culturally aligned robots are expected to be easily accepted by humans as part of daily life. In this paper, we aim at investigating the role of culture in representing robot emotions which are injected by humans during its early stage of development and subject to change through their own experience thereafter. Several public data sets of pictures labeled with affective ratings by Indian, American, and European subjects are presented to social humanoid Pepper robots. The result shows that robots can learn to behave socially in alignment with an individual's cultural background. Moreover, we have demonstrated that robots under the effect of different cultures can generate different behavioral responses to the same stimuli, which is considered one of the most important issues in socially assitive robotics.
Universal Dependencies (UD) is becoming a standard annotation scheme crosslinguistically, but it is argued that this scheme centering on content words is harder to parse than the conventional one centering on function words. To improve the parsability of UD, we propose a backand-forth conversion algorithm, in which we preprocess the training treebank to increase parsability, and reconvert the parser outputs to follow the UD scheme as a postprocess. We show that this technique consistently improves LAS across languages even with a state-of-the-art parser, in particular on core dependency arcs such as nominal modifier. We also provide an in-depth analysis to understand why our method increases parsability. 1
Recent work on the problem of latent tree learning has made it possible to train neural networks that learn to both parse a sentence and use the resulting parse to interpret the sentence, all without exposure to ground-truth parse trees at training time. Surprisingly, these models often perform better at sentence understanding tasks than models that use parse trees from conventional parsers. This paper aims to investigate what these latent tree learning models learn. We replicate two such models in a shared codebase and find that (i) only one of these models outperforms conventional tree-structured models on sentence classification, (ii) its parsing strategies are not especially consistent across random restarts, (iii) the parses it produces tend to be shallower than standard Penn Treebank (PTB) parses, and (iv) they do not resemble those of PTB or any other semantic or syntactic formalism that the authors are aware of.