Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
This article gives a survey of the main issues confronting the compilers of monolingual dictionaries in the age of the Internet. Among others, it discusses the relationship between a lexical database and a monolingual dictionary, the role of corpus evidence, historical principles in lexicography vs. synchronic principles, the instability of word meaning, the need for full vocabulary coverage, principles of definition writing, the role of dictionaries in society, and the need for dictionaries to give guidance on matters of disputed word usage. It concludes with some questions about the future of dictionary publishing. Keywords: Monolingual Dictionaries, Lexical Database, Dictionary Structure, Word Meaning, Meaning Change, Usage, Usage Notes, Historical Principles Of Lexicography, Synchronic Principles Of Lexicography, Register, Slang, Standard English, Vocabulary Coverage, Consistency Of Sets, Phraseology, Syntagmatic Patterns, Problems Of Compositionality, Linguistic Prescriptivism, Lexical Evidence **This article is an edited version of a plenary address delivered at the conference on 'Dictionaries, More Than Words', which took place at the Faculty of Social Sciences, University of Ljubljana, Ljubljana, Slovenia, 6 February 2009.
PURPOSE: To evaluate traumatized bone marrow with a dual-energy (DE) computed tomographic (CT) virtual noncalcium technique. MATERIALS AND METHODS: In this prospective institutional review board-approved study, 21 patients with an acute knee trauma underwent DE CT and magnetic resonance (MR) imaging. A software application was used to virtually subtract calcium from the images. Presence of fractures was noted, and presence of bone bruise was rated on a four-point scale for six femoral and tibial regions by two radiologists. CT numbers were obtained in the same regions. Consensus reading of independently read MR images served as the reference standard. Image ratings and CT numbers were subjected to receiver operating characteristic curve analysis. RESULTS: After exclusion of 16 regions owing to artifacts, MR imaging revealed 59 bone bruises in the remaining 236 regions (19 of 114 femoral, 40 of 122 tibial). Fractures were present in eight patients. Visual rating revealed areas under the curve of 0.886 and 0.897 in the femur and 0.974 and 0.953 in the tibia for observers 1 and 2, respectively. For CT numbers, the respective areas under the curve were 0.922 and 0.974. If scores of 1 and 2 (strong or mild bone bruise) were counted as positive, sensitivities were 86.4% and 86.4% and specificities were 94.4% and 95.5% for observers 1 and 2, respectively. The kappa statistic demonstrated good to excellent agreement (femur, kappa = 0.78; tibia, kappa = 0.87). CONCLUSION: This DE CT virtual noncalcium technique can subtract calcium from cancellous bone, allowing bone marrow assessment and potentially making posttraumatic bone bruises of the knee detectable with CT.
This paper explores whether and how a firm should adapt its strategy in view of consumer use of prior customer ratings. Specifically, we consider optimal pricing and whether the firm should offer an unexpected frill to early customers to enhance their product experiences. We show that if price history is unobserved by consumers, a forward-looking firm should always modify its strategy from single-period optimal one, but it may be optimal to do so by lowering price, by lowering price and offering frills, or by raising price and offering frills, depending on the market growth rate. Specifically, the last strategy becomes optimal when market growth rate is high enough. The results are similar when the price history is observed by consumers, except that no deviation from single-period profit maximization choices is optimal when market growth is low enough. We also analyze whether the firm should prefer that the price information be stated in or left out of consumer reviews. In addition, in considering the effects of consumer heterogeneity, we conclude that the optimal firm's effort to affect ratings is higher when the idiosyncratic part of consumer uncertainty is larger.
BACKGROUND: Recent advances have been made in the application of cognitive training strategies as interventions for mental disorders. One novel approach, cognitive control training (CCT), uses computer-based exercises to chronically increase prefrontal cortex recruitment. Activation of prefrontal control mechanisms have specifically been identified with attenuation of emotional responses. However, it is unclear whether recruitment of prefrontal resources alone is operative in this regard, or whether prefrontal control is important only in the role of explicit emotion regulation. This study examined whether exposure to cognitive tasks before an emotional challenge attenuated the effects of the emotional challenge. AIMS: We investigated whether a single training session could alter participants' reactivity to subsequent emotional stimuli on two computer-based tasks as well as affect ratings made during the study. We hypothesized that individuals performing the Cognitive Control (CC) task as compared to those performing the Peripheral Vision (PV) comparison task would (1) report reduced negative affect following the mood induction and the emotion task, and (2) exhibit reduced reactivity (defined by lower affective ratings) to negative stimuli during both the reactivity and recovery phases of the emotion task and (3) show a reduced bias towards threatening information. METHOD: Fifty-nine healthy participants were randomized to complete CC tasks or PV, underwent a negative mood induction, and then made valence and arousal ratings for IAPS images, and completed an assessment of attentional bias. RESULTS: RESULTS indicated that a single-session of CC did not consistently alter participants' responses to either task. However, performance on the CC tasks was correlated on subsequent ratings of emotional images. CONCLUSIONS: While overall these results do not support the idea that affective responding is altered by making healthy volunteers use their prefrontal cortex before the affective task, they are discussed in the context of study design issues and future research directions.
This paper shows that training a lexicalized parser on a lemmatized morphologically-rich treebank such as the French Treebank slightly improves parsing results. We also show that lemmatizing a similar in size subset of the English\nPenn Treebank has almost no effect on parsing performance with gold lemmas and leads to a small drop of performance when automatically assigned lemmas and POS tags are used. This highlights two facts: (i) lemmatization helps to reduce lexicon data-sparseness issues for French, (ii) it also makes the parsing process sensitive to correct assignment of POS tags to unknown words.
Co-constructing communicative effectiveness is often challenging in English as a lingua franca (ELF): speakers have considerably less to go on in terms of shared expectations of cultural knowledge and linguistic norms. A university environment provides a convenient backdrop for sharing at least academic conventions – although these vary more than might be surmised from the uniform labelling of such event types. This paper looks into some discourse and lexicogrammatical features in academic ELF, using ELFA as the database. The data consists of spoken language, which provides direct access to the ways in which meanings are negotiated in ongoing discourse, and the speech events are typically polylogic. ELF discourse requires close cooperation from the participants, which is reflected in its enhanced explicitness among other things. The explicitation strategies speakers display facilitate mutual comprehensibility and contribute to social cohesion within the multi-participant groups. Such strategies also help overcome the potential problems participants might have in dealing with a variety of formal deviations from ordinary English as a native language (ENL). Most of the time ELF bears a very close resemblance to Standard English, but signs of incipient ELF-specific developments are also in evidence.
A commonly held assumption is that processes underlying explicit and implicit memory are distinct. Recent evidence, however, suggests that they may interact more than previously believed. Using the remember-know procedure the current study examines the relation between recollection, a process thought to be exclusive to explicit memory, and performance on two implicit memory tasks, lexical decision and word stem completion. We found that, for both implicit tasks, words that were recollected were associated with greater priming effects than were words given a subsequent familiarity rating or words that had been studied but were not recognised (misses). Broadly, our results suggest that non-voluntary processes underlying explicit memory also benefit priming, a measure of implicit memory. More specifically, given that this benefit was due to a particular aspect of explicit memory (recollection), these results are consistent with some strength models of memory and with Moscovitch's (2008) proposal that recollection is a two-stage process, one rapid and unconscious and the other more effortful and conscious.
Ott, N. & R. Ziai (2010). Evaluating dependency parsing performance on german learner language. In M. Dickinson, K. Müürisep & M. Passarotti (eds.), Proceedings of the Ninth International Workshop on Treebanks and Linguistic Theories. Vol. 9 of NEALT Proceeding Series, 175–186.
We show that the standard beam-search algorithm can be used as an efficient decoder for the global linear model of Zhang and Clark (2008) for joint word segmentation and POS-tagging, achieving a significant speed improvement. Such decoding is enabled by: (1) separating full word features from partial word features so that feature templates can be instantiated incrementally, according to whether the current character is separated or appended; (2) deciding the POS-tag of a potential word when its first character is processed. Early-update is used with perceptron training so that the linear model gives a high score to a correct partial candidate as well as a full output. Effective scoring of partial structures allows the decoder to give high accuracy with a small beam-size of 16. In our 10-fold crossvalidation experiments with the Chinese Treebank, our system performed over 10 times as fast as Zhang and Clark (2008) with little accuracy loss. The accuracy of our system on the standard CTB 5 test was competitive with the best in the literature. 1
u-tokyo.ac.jp Several recent discourse parsers have employed fully-supervised machine learning approaches. These methods require human annotators to beforehand create an extensive training corpus, which is a time-consuming and costly process. On the other hand, unlabeled data is abundant and cheap to collect. In this paper, we propose a novel semi-supervised method for discourse relation classification based on the analysis of cooccurring features in unlabeled data, which is then taken into account for extending the feature vectors given to a classifier. Our experimental results on the RST Discourse Treebank corpus and Penn Discourse Treebank indicate that the proposed method brings a significant improvement in classification accuracy and macro-average F-score when small training datasets are used. For instance, with training sets of c.a. 1000 labeled instances, the proposed method brings improvements in accuracy and macro-average F-score up to 50% compared to a baseline classifier. We believe that the proposed method is a first step towards detecting low-occurrence relations, which is useful for domains with a lack of annotated data. 1
This article details a series of carefully designed experiments aiming at evaluating the influence of automatic pre-annotation on the manual part-of-speech annotation of a corpus, both from the quality and the time points of view, with a specific attention drawn to biases. For this purpose, we manually annotated parts of the Penn Treebank corpus (Marcus et al., 1993) under various experimental setups, either from scratch or using various pre-annotations. These experiments confirm and detail the gain in quality observed before (Marcus et al., 1993; Dandapat et al., 2009; Rehbein et al., 2009), while showing that biases do appear and should be taken into account. They finally demonstrate that even a not so accurate tagger can help improving annotation speed. 1
The contextual cueing effect (CC) refers to the phenomenon in which visual search performance is faster for targets appearing in previously exposed configurations than for targets appearing in new configurations. We investigated whether the learned configurations affect other types of response such as affective evaluation. Participants were asked to search T-target among rotated L-distractors. The mean reaction time showed a typical CC. Then, participants were asked to evaluate how much they like the repeated or new configurations (Experiment 1), or how much they feel the difficulty of target detection (Experiment 2). The results showed that both the liking and difficulty ratings for the repeated configurations were lower than those for the new configurations.
This paper proposes a unified framework for zero anaphora resolution, which can be divided into three sub-tasks: zero anaphor detection, anaphoricity determination and antecedent identification. In particular, all the three sub-tasks are addressed using tree kernel-based methods with appropriate syntactic parse tree structures. Experimental results on a Chinese zero anaphora corpus show that the proposed tree kernel-based methods significantly outperform the feature-based ones. This indicates the critical role of the structural information in zero anaphora resolution and the necessity of tree kernel-based methods in modeling such structural information. To our best knowledge, this is the first systematic work dealing with all the three sub-tasks in Chinese zero anaphora resolution via a unified framework. Moreover, we release a Chinese zero anaphora corpus of 100 documents, which adds a layer of annotation to the manually-parsed sentences in the Chinese Treebank (CTB) 6.0. 1
Inducing a grammar from text has proven to be a notoriously challenging learning task despite decades of research. The primary reason for its difficulty is that in order to induce plausible grammars, the underlying model must be capable of representing the intricacies of language while also ensuring that it can be readily learned from data. The majority of existing work on grammar induction has favoured model simplicity (and thus learnability) over representational capacity by using context free grammars and first order dependency grammars, which are not sufficiently expressive to model many common linguistic constructions. We propose a novel compromise by inferring a probabilistic tree substitution grammar, a formalism which allows for arbitrarily large tree fragments and thereby better represent complex linguistic structures. To limit the model’s complexity we employ a Bayesian non-parametric prior which biases the model towards a sparse grammar with shallow productions. We demonstrate the model’s efficacy on supervised phrase-structure parsing, where we induce a latent segmentation of the training treebank, and on unsupervised dependency grammar induction. In both cases the model uncovers interesting latent linguistic structures while producing competitive results. Keywords: grammar induction, tree substitution grammar, Bayesian non-parametrics, Pitman-Yor process, Chinese restaurant process 1.
Computer-asisted language learning needs better lexical databases in order to produce better software for vocabulary learning. This paper attempts to give some guidelines for the construction of a dedicated lexical database for vocabulary learning purposes.
Traditional dictionaries provide lexicographic data of the same type in different ways; such unrestricted freedom might result in imprecise information. Some such cases are explained by an attempt to compile traditional dictionaries so that they are handy for their users. The paper discusses several non-systemic cases of providing forms of lexicographic items, their meanings and shades of meaning, derivatives, proper and common lexis. The data has been drawn from more than ten dictionaries of several languages. The structure of the electronic database requires utmost precision; therefore, any non-systemic items are not justified. Due to the above reasons, a lexeme, or a lexicographic unit in one of its meanings, has been chosen as the key structural unit of the Database of Contemporary Lithuanian Lexicon (DCLL, Dabartinės lietuvių leksikos duomenų bazė). To produce a most precise card of a lexeme is much more rational and simpler than that of the whole lexicographic unit. In the process, a number of analogous cases and other cases of non-systemic description of data should be taken into account. Moreover, it is hardly possible to consider them all. Thus an attempt will be made to finally produce the DCLL as a card index of lexemes rather than lexicographic units (their entries). Dictionary entries are to be produced by joining the lexemes automatically.
This study examined whether breathing rate affected self-reported pain and emotion following thermal pain stimuli in women with fibromyalgia syndrome (FM: n=27) or age-matched healthy control women (HC: n=25). FM and HC were exposed to low and moderate thermal pain pulses during paced breathing at their normal rate and one-half their normal rate. Thermal pain pulses were presented in four blocks of four trials. Each block included exposure to both mild and moderate pain trials, and periods of both normal and slow paced breathing. Pain intensity and unpleasantness were recorded immediately following each pain trial, and positive and negative affect were assessed at the end of each block of trials. Compared to normal breathing, slow breathing reduced ratings of pain intensity and unpleasantness, particularly for moderately versus mildly painful thermal stimuli. The effects of slow breathing on pain ratings were less reliable for FM patients than for HCs. Slow versus normal breathing decreased negative affect ratings following thermal pain pulses for both groups, and increased positive affect reports, but only for healthy controls with high trait negative affect. Participants who reported higher levels of trait positive affect prior to the experiment showed greater decreases in negative affect as a result of slow versus normal breathing. These experimental findings provide support for prior reports on the benefits of yogic breathing and mindful Zen meditation for pain and depressed affect. However, chronic pain patients may require more guidance to obtain therapeutic benefit from reduced breathing rates.
In the architecture of a natural language processing system based on linguistic knowledge, two types of component are important: the knowledge databases and the processing modules. One of the knowledge databases is the lexical database, which is responsible for providing the lexical unities and its properties to the processing modules. The systems that process two or more languages require bilingual and/or multilingual lexical databases. These databases can be constructed by aligning distinct monolingual databases. In this paper, we present the interlingua and the strategy of aligning the two monolingual databases in REBECA, which only stores concepts from the “wheeled vehicle” domain.
We describe a study that examined the distinctiveness of momentary positive and negative affect (PA and NA) and their relationships with blood pressure and heart rate at work in a sample of 67 full-time employees who provided psychological and physiological measurements multiple times a day, over 10 working days. These employees recorded their subjective ratings on personal digital assistant devices programmed specifically for this study, and assessed their blood pressure and heart rate using automatic recording devices, at each measurement period. Results support the distinctiveness of PA and NA. First, within-individual factor analyses confirmed the existence of two underlying factors in the momentary affect ratings. Second, the within-individual correlations between PA and NA scores (-.14), and among the two latent constructs (-.13) were both low in magnitude. Finally, as predicted by theory, NA was significantly related to blood pressure within individuals, whereas PA and blood pressure were unrelated (both PA and NA were related to heart rate within individuals, as predicted).
Accelerated growth of the Internet has enabled users worldwide to share their feelings and experiences. User-generated content (UGC) websites are the most abundant sources of user reviews. Accurately identifying sentiment phrases is essential to understand the expressed opinions in user reviews. To achieve this, part-of-speech (POS) patterns of phrases are useful. However, previous studies for Chinese opinion extraction only translate English POS patterns directly into Chinese for this task without considering the feasibility. Therefore, this work proposes a Chinese opinion extraction method that exploits the observed Sinica Treebank POS patterns for sentiment phrase identification. Sinica Treebank is a widely representative POS corpus for Chinese. The results of preliminary experiments indicate that the proposed method is highly effective in extracting opinions from Chinese UGC reviews.
In this paper we present a novel phrase structure parsing approach with the help of dependency structure. Different with existing phrase parsers, in our approach the inference procedure is guided by dependency structure, which makes the parsing procedure flexibly. The experimental results show our approach is much more accurate. With the help of golden dependency trees, F1 score of our parser achieves 96.08 % on Penn English Treebank and 90.61 % on Penn Chinese Treebank. With the help of N-best dependency trees generated by modified
The present study investigates whether dementia patients can make accurate social and affective judgments of face stimuli. We used verbal material as a control condition to see whether dementia patients are capable of using the rating scales the same way as controls. We compared the trustworthiness, valence, and arousal ratings of dementia patients with those from a healthy control group. Participants rated pictures of young and old, female and male neutral faces and two fictitious biographies. The results indicate that dementia patients make accurate judgments of unfamiliar faces with respect to trustworthiness and valence. Results concerning arousal ratings suggest that the corresponding scale might be difficult to use for dementia patients and possibly for older participants as well.
The field of lexicography is undergoing a major revolution. The rapid replacement of the traditional paper dictionary by electronic dictionaries opens up exciting possibilities but also constitutes a major challenge to the field. The 'eLexicography in the 21st Century: New Challenges, New Applications' conference organized by the Centre for English Corpus Linguistics of the Université catholique de Louvain in October 2009 aimed to bring together the many researchers around the world who are working in the fast developing field of electronic lexicography and to act as a showcase for the latest lexicographic developments and software solutions in the field. The conference attracted both academics and industrial partners from 30 different countries who presented electronic dictionary projects dealing with no less than 22 languages.\nThe resulting proceedings volume bears witness to the tremendous vitality and diversity of research in the field. The volume covers a wide range of topics, including:\n- the use of language resources for lexicographic purposes, in the form of lexical databases like WordNet or corpora of different types\n- innovative changes to the dictionary structure afforded by the electronic medium, in particular multiple access routes and efficient integration of phraseology\n- specialised dictionaries (e.g. SMS dictionaries, sign language dictionaries)\n- automated customisation of dictionaries in function of users' needs\n- exploitation of Natural Language Processing tools\n- integration of electronic dictionaries into language learning and teaching tools
This paper explores ways to detect errors in aligned corpora, using very little technology. In the first method, applicable to any aligned corpus, we consider alignment as a string-to-string mapping. Treating the target string as a label, we examine each source string to find inconsistencies in alignment. Despite setting up the problem on a par with grammatical annotation, we demonstrate crucial differences in sorting errors from legitimate variations. The second method examines phrase nodes which are predicted to be aligned, based on the alignment of their yields. Both methods are effective in complementary ways. 1
We present an approach to automatically identifying the arguments of discourse connectives based on data from the Penn Discourse Treebank. Of the two arguments of connectives, called Arg1 and Arg2, we focus on Arg1, which has proven more challenging to identify. Our approach employs a sentence-based representation of arguments, and distinguishes intra-sentential connectives, which take both their arguments in the same sentence, from inter-sentential connectives, whose arguments are found in different sentences. The latter are further distinguished by paragraph position into ParaInit connectives, which appear in a paragraph-initial sentence, and ParaNonInit connectives, which appear elsewhere. The paper focusses on predicting Arg1 of Inter-sentential ParaNonInit connectives, presenting a set of scope-based filters that reduce the search space for Arg1 from all the previous sentences in the paragraph to a subset of them. For cases where these filters do not uniquely identify Arg1, coreference-based heuristics are employed. Our analysis shows an absolute 3 % performance improvement over the high baseline of 83.3 % for identifying Arg1 of Inter-sentential ParaNonInit connectives. 1.
Abstract The aim of the article is to introduce a new approach to verb valency analysis. This approach – full valency – observes properties of verbs which occur solely in actual language usage. The term “full valency” means that all arguments, without distinguishing complements (obligatory arguments governed by the verb) and adjuncts (optional arguments directly dependent on the predicate verb), are taken into account. Because of an expectation that full valency reflects some mechanism which governs verb behaviour in a language, hypotheses concerning (1) the distribution of full valency frames, (2) the relationship between the number of valency frames and the frequency of the verb, and (3) the relationship between the number of valency frames and verb length were tested empirically. To test the hypotheses, a Czech syntactically annotated corpus – the Prague Dependency Treebank – was used.
OBJECTIVES: It is well recognised that medical training can be extremely stressful and that high stress is a risk factor for a wide range of psychological and health-related consequences. The primary aims of this study were to introduce the Medical Student Stress Profile (MSSP) and to demonstrate its psychometric quality as a specific device for auditing medical student stress. Secondary aims were to establish the reliability, construct and criterion validity of this instrument and to explore the relationships between stress, coping, personality, motivation and emotional intelligence in medical students. METHODS: A battery of self-report measures including the MSSP was administered to a sample of 239 undergraduate and graduate-entry medical students. The battery included indices of stress, coping with and proneness to stress, as well as measures of emotional intelligence, motivation style, personality traits, educational environment perception and self-reported symptomatology. Psychometric evaluation of the MSSP was conducted along with a correlation analysis of stress concomitants. RESULTS: The MSSP revealed good psychometric properties and showed a substantial stress load in the participant sample. The pattern of correlations with concomitant measures conformed generally to expectations. Strong cohort effects were observed, which suggest the importance of future investigation into the role of the group in stress amelioration. Stress adversely affects ratings of the educational environment as measured by the Dundee Ready Education Environment Measure. CONCLUSIONS: The MSSP was specifically developed for the medical training context and may have utility for individual and group stress audits of medical students and as a device to inform remedial programmes in stress management in medical education.
In the field of natural language processing (NLP), there often exist multiple corpora with different annotation standards for the same task. In this paper, we take syntactic parsing as a case study and propose a reranking method which is able to make direct use of disparate treebanks simultaneously without using techniques such as treebank conversion. The method proceeds in three steps: 1) build parsers on individual treebanks; 2) use parsers independently to generate n-best lists for each sentence in test set; 3) rerank individual n-best lists which correspond to the same sentence by using consensus information exchanged among these n-best lists. Experimental results on two open Chinese treebanks show that our method significantly outperforms the baseline system by 0.84% and 0.53% respectively.
Problem statement: Text summarization takes care of choosing the most significant portions of text and generates coherent summaries that express the main intent of the given document. This study aims to compare the performances of the three text summarization systems developed by the authors with some of the existing Summarization systems available. These three approaches to text summarization are based on semantic nets, fuzzy logic and evolutionary programming respectively. All the three represent approaches to achieve connectionism. Approach: First approach performs Part of Speech (POS) tagging, semantic and pragmatic analysis and cohesion. The second system under discussion was a new extraction based automated system for text summarization using a decision module that employs fuzzy concepts. Third system under consideration was based on a combination of evolutionary, fuzzy and connectionist techniques. Results: Semantic net approach performs better than the MS Word summarizer as far as the semantics of the original text was concerned. To compare our summaries with those of the well known MS Word, Intellexer and Copernic summarizers, we use DUC's human generated summaries as the bench-mark. The results were very encouraging. The second approach based on fuzzy logic results in an efficient system since fuzzy logic mimics decision making of humans. Third system showed promising results as far as precision and F-measure are concerned than all the other approaches. Conclusion: Our first approach used WordNet, a lexical database for English. Unlike other dictionaries, WordNet does not include information about etymology, pronunciation and the forms of irregular verbs and contains only limited information about usage. To overcome this limitation, we developed a new text summarizer based on fuzzy logic. As Text summarization application requires learning ability based on activation, we utilize ANN attribute through a connectionist model to achieve the best results.
On October 8‐10, 2009, an interdisciplinary group met in Beverley, Massachusetts, to evaluate the state of the art in the computational modeling of narrative. Three important findings emerged: (1) current work in computational modeling is described by three different levels of representation; (2) there is a paucity of studies at the highest, most abstract level aimed at inferring the meaning or message of the narrative; and (3) there is a need to establish a standard data bank of annotated narratives, analogous to the Penn Treebank.
We propose to analyse semantic similarity in comparable text by matching syntactic trees and labeling the alignments according to one of five semantic similarity relations. We present a Memory-based Graph Matcher (MBGM) that performs both tasks simultaneously as a combination of exhaustive pairwise classification using a memory-based learner, followed by global optimization of the alignments using a combinatorial optimization algorithm. The method is evaluated on a monolingual treebank consisting of comparable Dutch news texts. Results show that it performs substantially above the baseline and close to the human reference.
In the Chinese teaching and research,it often needs to draw syntax tree for analyzing of relationship among compositions of sentence.Drawn syntax tree manually has many defects,such as having huge workload,needing immense storage capacity and so on.This paper proposes a method to generate syntax tree automatically and display the syntax tree in web page by using VML technology.By comparing the syntax trees generated in this paper with the syntax trees from Treebank of PKU,the result shows that the method of coordinate computing proposed in this paper is more precision,which also demonstrates the efficiency of the visualization method.
In this paper, we present what we believe to be the first data-driven dependency parser for Urdu. The parser was trained and tuned using MaltParser system, a system for data-driven dependency parsing. The Urdu dependency treebank (UDT) is used for training and testing of the Urdu dependency parser, is also presented first time. The UDT contains corpus of 2853 sentences which are annotated at multiple levels such as part-of-speech (POS) level, chunk (phrase level) and dependency relations level. The UDT also contains information about the token counter, head of current token. The annotation is done manually to build UDT. Urdu Dependency Parsing system is evaluated by conducting a series of experiments. All experiments are performed using Maltparser default algorithm with different feature models. Initial, a base line simple feature model consisting word position, word, head and dependency relation is used for Urdu dependency parsing. Then feature model is enhanced by adding part-of-speech (POS) and chunk (Phrase level) information. The results of all parsing experiments are reported. The overall best labeled accuracy (LA) achieved 74.48 % and 90.14 % of unlabeled attachment score (UAS) is achieved. The error analysis is performed by comparing output data with treebank test data which manual parsed to analyze and classify the different types of errors produced by the parser. This is very useful to identify the future directions for future expansion of the treebank and for improving the parsing accuracy. 1.
Current automatic wrappers using DOM tree and visual properties of data records to extract the required information from the deep web generally have limitations such as the inability to check the similarity of tree structures accurately. Our study shows that data records located in the deep web do not only share similar visual properties and tree structures, but they are also related semantically in their contents. As such we are able to propose an ontological technique using existing lexical database for English (WordNet) for the extraction of data records from deep web pages. Wrappers designed based on ontological technique are able to reduce the number of potential data regions identified for data extraction, thus improve the data extraction accuracy. In this study, we use visual cue from the underlying browser rendering engine to locate and extract the relevant data region from the deep web by measuring the text and image sizes of data records. Experimental results show that our technique is robust and performs better than the existing state of the art wrappers. Unlike existing ontological based wrappers, our wrapper is domain independent and is able to extract wide range of data records with different structures.
We present the first evaluation of the utility of automatic evaluation metrics on surface realizations of Penn Treebank data. Using outputs of the OpenCCG and XLE realizers, along with ranked WordNet synonym substitutions, we collected a corpus of generated surface realizations. These outputs were then rated and post-edited by human annotators. We evaluated the realizations using seven automatic metrics, and analyzed correlations obtained between the human judgments and the automatic scores. In contrast to previous NLG meta-evaluations, we find that several of the metrics correlate moderately well with human judgments of both adequacy and fluency, with the TER family performing best overall. We also find that all of the metrics correctly predict more than half of the significant systemlevel differences, though none are correct in all cases. We conclude with a discussion of the implications for the utility of such metrics in evaluating generation in the presence of variation. A further result of our research is a corpus of post-edited realizations, which will be made available to the research community. 1
This paper proposes a dependency parsing method that uses bilingual constraints to improve the accuracy of parsing bilingual texts (bitexts). In our method, a targetside tree fragment that corresponds to a source-side tree fragment is identified via word alignment and mapping rules that are automatically learned. Then it is verified by checking the subtree list that is collected from large scale automatically parsed data on the target side. Our method, thus, requires gold standard trees only on the source side of a bilingual corpus in the training phase, unlike the joint parsing model, which requires gold standard trees on the both sides. Compared to the reordering constraint model, which requires the same training data as ours, our method achieved higher accuracy because of richer bilingual constraints. Experiments on the translated portion of the Chinese Treebank show that our system outperforms monolingual parsers by 2.93 points for Chinese and 1.64 points for English. 1
RATIONALE: Acute tryptophan depletion (ATD) decreases levels of central serotonin. ATD thus enables the cognitive effects of serotonin to be studied, with implications for the understanding of psychiatric conditions, including depression. OBJECTIVE: To determine the role of serotonin in conscious (explicit) and unconscious/incidental processing of emotional information. MATERIALS AND METHODS: A randomized, double-blind, cross-over design was used with 15 healthy female participants. Subjective mood was recorded at baseline and after 4 h, when participants performed an explicit emotional face processing task, and a task eliciting unconscious processing of emotionally aversive and neutral images presented subliminally using backward masking. RESULTS: ATD was associated with a robust reduction in plasma tryptophan at 4 h but had no effect on mood or autonomic physiology. ATD was associated with significantly lower attractiveness ratings for happy faces and attenuation of intensity/arousal ratings of angry faces. ATD also reduced overall reaction times on the unconscious perception task, but there was no interaction with emotional content of masked stimuli. ATD did not affect breakthrough perception (accuracy in identification) of masked images. CONCLUSIONS: ATD attenuates the attractiveness of positive faces and the negative intensity of threatening faces, suggesting that serotonin contributes specifically to the appraisal of the social salience of both positive and negative salient social emotional cues. We found no evidence that serotonin affects unconscious processing of negative emotional stimuli. These novel findings implicate serotonin in conscious aspects of active social and behavioural engagement and extend knowledge regarding the effects of ATD on emotional perception.
Tree-based translation models, which exploit the linguistic syntax of source language, usually separate decoding into two steps: parsing and translation. Although this separation makes tree-based decoding simple and efficient, its translation performance is usually limited by the number of parse trees offered by parser. Alternatively, we propose to parse and translate jointly by casting tree-based translation as parsing. Given a source-language sentence, our joint decoder produces a parse tree on the source side and a translation on the target side simultaneously. By combining translation and parsing models in a discriminative framework, our approach significantly outperforms a forestbased tree-to-string system by 1.1 absolute BLEU points on the NIST 2005 Chinese-English test set. As a parser, our joint decoder achieves an F1 score of 80.6 % on the Penn Chinese Treebank. 1
Facial electromyography (EMG) is a useful physiological measure for detecting subtle affective changes in real time. A time series of EMG data contains bursts of electrical activity that increase in magnitude when the pertinent facial muscles are activated. Whereas previous methods for detecting EMG activation are often based on deterministic or externally imposed thresholds, we used regime-switching models to probabilistically classify each individual’s time series into latent “regimes” characterized by similar error variance and dynamic patterns. We also allowed the association between EMG signals and self-reported affect ratings to vary between regimes and found that the relationship between these two markers did in fact vary over time. The potential utility of using regime-switching models to detect activation patterns in EMG data and to summarize the temporal characteristics of EMG activities is discussed.
Much of the work on Statistical Machine Translation (SMT) from morphologically rich languages has shown that morphological tokenization and orthographic normalization help improve SMT quality because of the sparsity reduction they contribute. In this paper, we study the effect of these processes on SMT when translating into a morphologically rich language, namely Arabic.We explore a space of tokenization schemes and normalization options. We only evaluate on detokenized and orthographically correct (enriched) output. Our results show that the best performing tokenization scheme is that of the Penn Arabic Treebank. Additionally, training on orthographically normalized (reduced) text then jointly enriching and detokenizing the output outperforms training on enriched text.