Papers reviewed and determined not to be word norm studies. Use the flag icon to report errors or suggest re-inclusion.
18625 papers
This paper discusses two new procedures for extracting verb valences from raw texts, with an application to the Polish language. The first novel technique, the EM selection algorithm, performs unsupervised disambiguation of valence frame forests, obtained by applying a non-probabilistic deep grammar parser and some post-processing to the text. The second new idea concerns filtering of incorrect frames detected in the parsed text and is motivated by an observation that verbs which take similar arguments tend to have similar frames. This phenomenon is described in terms of newly introduced co-occurrence matrices. Using co-occurrence matrices, we split filtering into two steps. The list of valid arguments is first determined for each verb, whereas the pattern according to which the arguments are combined into frames is computed in the following stage. Our best extracted dictionary reaches an F-score of 45%, compared to an F-score of 39% for the standard frame-based BHT filtering.
An impressive amount of work was devoted over the past few decades to collocation extraction. The state of the art shows that there is a sustained interest in the morphosyntactic preprocessing of texts in order to better identify candidate expressions; however, the treatment performed is, in most cases, limited (lemmatization, POS-tagging, or shallow parsing). This article presents a collocation extraction system based on the full parsing of source corpora, which supports four languages: English, French, Spanish, and Italian. The performance of the system is compared against that of the standard mobile-window method. The evaluation experiment investigates several levels of the significance lists, uses a fine-grained annotation schema, and covers all the languages supported. Consistent results were obtained for these languages: parsing, even if imperfect, leads to a significant improvement in the quality of results, in terms of collocational precision (between 16.4 and 29.7%, depending on the language; 20.1% overall), MWE precision (between 19.9 and 35.8%; 26.1% overall), and grammatical precision (between 47.3 and 67.4%; 55.6% overall). This positive result bears a high importance, especially in the perspective of the subsequent integration of extraction results in other NLP applications.
The embodied cognition hypothesis suggests that motor and premotor areas are automatically and necessarily involved in understanding action language, as word conceptual representations are embodied. This transcranial magnetic stimulation (TMS) study explores the role of the left primary motor cortex in action-verb processing. TMS-induced motor-evoked potentials from right-hand muscles were recorded as a measure of M1 activity, while participants were asked either to judge explicitly whether a verb was action-related (semantic task) or to decide on the number of syllables in a verb (syllabic task). TMS was applied in three different experiments at 170, 350 and 500 ms post-stimulus during both tasks to identify when the enhancement of M1 activity occurred during word processing. The delays between stimulus onset and magnetic stimulation were consistent with electrophysiological studies, suggesting that word recognition can be differentiated into early (within 200 ms) and late (within 40)
The aim of the paper is to show that a subset of Text Encoding Initiative Guidelines is a reasonable choice as a standard for stand-off XML encoding of syntactically annotated corpora. The proposed TEI schema — actually employed in the National Corpus of Polish — is compared to other such candidate standards, including TIGER-XML, SynAF and PAULA. 1
Four experiments employed a priming methodology to investigate different mechanisms of stress assignment and how they are modulated by lexical and sub-lexical mechanisms in reading aloud in Italian. Lexical stress is unpredictable in Italian, and requires lexical look-up. The most frequent stress pattern (Dominant) is on the penultimate syllable [laVOro (work)], while stress on the antepenultimate syllable [MAcchina (car)] is relatively less frequent (non-Dominant). Word and pseudoword naming responses primed by words with non-dominant stress - which require whole-word knowledge to be read correctly - were compared to those primed by nonwords. Percentage of errors to words and percentage of dominant stress responses to nonwords were measured. In Experiments 1 and 2 stress errors increased for non-dominant stress words primed by nonwords, as compared to when they were primed by words. The results could be attributed to greater activation of sub-lexical codes, and an associated tendency)
WordNet, a lexical database for English, is organized around semantic and lexical relationships between synsets, concepts represented by sets of synonymous word senses. Offering reasonably comprehensive coverage of the nouns, verbs, adjectives, and adverbs of general English, WordNet is a widely used resource for dealing with the ambiguity that arises from homonymy, polysemy, and synonymy. WordNet is used in many information-related tasks and applications (e.g., word sense disambiguation, semantic similarity, lexical chaining, alignment of parallel corpora, text segmentation, sentiment and subjectivity analysis, text classification, information retrieval, text summarization, question answering, information extraction, and machine translation).
From the margins of discourse, this autobiographical work disrupts the traditional HEGEMONIC narratives of THEIR INSTITUTIONS and challenges the conventional notions of the genre of autobiography through its form and content. It uncovers and upsets both WESTERN and Punjabi PATRIArchal definitions of THEIR/Their CULTure: Furthermore, it undermines linguistic norms by using a mix of English, Punjabi and some French. Through narratives, prose, poetry, transparent language/languages my words break the long kept silences of three Punjabi women living amidst racism, sexism and classism in THEIR/Their SOCIety Public and private History of THEIR/Their CULTure, fiction and fantasy are juxtaposed through mixed genres and mixed codes, and the fragmented units are not necessarily presented as chronologically linear. Unlike typical autobiographies, I braid together the restrictive lives and painful experiences of my mother, my grandmother, and myself in an attempt to show how THEIR/Their oppressive History and CULTure shape and have shaped our life/lives, our acceptance and our resistance. In doing so I begin to see parallels between the oppression of women and untouchables in Indian Society and the exploitation of minority women and minorities in CANADIAN SOCIETY. My experiment with structure and typography helps me, a semi-trilingual, working/middle class minority woman and EDUCAtor, to construct and deconstruct my personal self and selves in a more meaningful way. This type of meaning making takes precedence over conventional stylistic or established structures. My identity cannot be separated from the languages, CULTures, in which I wander endlessly or from the oppressive conditions of those whose lives have shaped and continue to shape me. Despite the pressure, I cannot choose between these identities. As a result I am forced to continue to negotiate between my identities and live an intricate web of shifting power relations. I realise that the non-contradictory self is impossible.Therefore I must go beyond and create something new. Something that is more than the mere bringing together of identities ~ it is something in the beyond, kithe uther ~ gahan. From my third space I call upon minority women, white women, EDUCators, readers and writers to also go beyond and find a THIRD Space from which to work together in order to reposition the positions of POWER that have been placed upon us through centuries of Oppression and EXPLOITATION. Only by speaking out in my languages, my words, my experiences, my poet's voice, my storying voice, my minority woman's voice, my TEACHer's voice, my fragmented voices, will I begin to overcome the tradition of silence and continue the necessary resistance.
We present an approach for smoothing treebank-PCFG lexicons by interpolating treebank lexical parameter estimates with estimates obtained from unannotated data via the Inside-outside algorithm. The PCFG has complex lexical categories, making relative-frequency estimates from a treebank very sparse. This kind of smoothing for complex lexical categories results in improved parsing performance, with a particular advantage in identifying obligatory arguments subcategorized by verbs unseen in the treebank.
In the context of debates on global languages, this chapter takes a ‘top-down’ perspective in exploring the emergence of current panhispanic language policies, led by the Real Academia Espanola (RAE) and involving close collaboration with, amongst others, the Instituto Cervantes. The relationship between these two organizations has been depicted by a former RAE Director as “ministros de una misma iglesia; nosotros somos los padres conciliares y [el Instituto Cervantes] los misioneros” (Garcia de la Concha 2005:5). We argue that these policies shape, and are shaped by, widespread language ideological debates about standardization and language unity across the cohort of Spanish language academies and amongst language commentators in Spain, Latin America, and the U.S. We examine three facets of language policies currently being pursued by the Spanish government which we term: 1) the ‘internal’ policies, 2) the ‘external’ policies, and, of increasing importance and the central focus of this chapter, 3) the ‘panhispanic language’ policies. We critically analyze these policies and language planning practices, seeking to identify the agents that are responsible for enacting and implementing them. Of particular interest is the series of Congresos de la Lengua Espanola which plays an important role in policy development and promotion. This chapter considers how institutional language ideologies drive Spanish language spread in the global context of a linguistic ‘market’, noting that private business interests in Spain have awoken to the huge commercial opportunities offered by the Spanish-speaking markets and the selling of the Spanish language. We suggest that panhispanism has become synonymous for the RAE with globalization and ask what the consequences of this might be in situations of language contact with Iberian and American languages.
In this paper, we address the problem of document re-ranking in information retrieval, which is usually conducted after initial retrieval to improve rankings of relevant documents. To deal with this problem, we propose a method which automatically constructs a term resource specific to the document collection and then applies the resource to document re-ranking. The term resource includes a list of terms extracted from the documents as well as their weighting and correlations computed after initial retrieval. The term weighting based on local and global distribution ensures the re-ranking not sensitive to different choices of pseudo relevance, while the term correlation helps avoid any bias to certain specific concept embedded in queries. Experiments with NTCIR3 data show that the approach can not only improve performance of initial retrieval, but also make significant contribution to standard query expansion.
Grammar extraction in deep formalisms has received remarkable attention in recent years. We recognise its value, but try to create a more precision-oriented grammar, by hand-crafting a core grammar, and learning lexical types and lexical items from a treebank. The study we performed focused on German, and we used the Tiger treebank as our resource. A completely hand-written grammar in the framework of HPSG forms the inspiration for our core grammar, and is also our frame of reference for evaluation.
An open problem in dependency parsing is the accurate and efficient treatment of non-projective structures. We propose to attack this problem using chart-parsing algorithms developed for mildly context-sensitive grammar formalisms. In this paper, we provide two key tools for this approach. First, we show how to reduce non-projective dependency parsing to parsing with Linear Context-Free Rewriting Systems (LCFRS), by presenting a technique for extracting LCFRS from dependency treebanks. For efficient parsing, the extracted grammars need to be transformed in order to minimize the number of nonterminal symbols per production. Our second contribution is an algorithm that computes this transformation for a large, empirically relevant class of grammars.
The rise of a standard language is inextricably connected to value judgements about linguistic variants. During standardisation processes certain linguistic expressions are marked as ‘correct’ and prestigious and subsequently selected as a standard form whereas other linguistic features are labelled as ‘bad’ and corrupt use of language. As Haugen rightly points out, ‘[w]here a norm is to be established, the problem will be as complex as the sociolinguistic structure of the people involved’ (1997, p. 349). It is, after all, the socio-political context that influences the evaluation of the language usage. An established norm, in turn, has many socio-political consequences. This work will be concerned with the establishment of linguistic norms in the history of specific languages and the socio-political contexts in which these norms arose as well as their influence on actual usage. More precisely, this study seeks to trace the development of the subjunctive mood in English and German, with a special focus on the Austrian variety,1 during part of their standardisation processes, namely the eighteenth century. As grammarians were attempting to shape and codify a prestige variety during this period, the question arises whether and to what extent these normative grammarians influenced the development of the inflectional subjunctive. After all, the subjunctive mood has been claimed to have been on the decline in both English and German in the eighteenth century (cf. for English: Strang, 1970, p. 209; Turner, 1980, p. 272; Görlach, 2001, p. 122; for German: von Polenz, 1994, pp. 261–263).
The Columbia Arabic Treebank (CATiB) is a database of syntactic analyses of Arabic sentences. CATiB contrasts with previous ap-proaches to Arabic treebanking in its emphasis on faster production with some constraints on linguistic richness. Two basic ideas inspire the CATiB approach. First, CATiB avoids the annotation of redundant linguistic information that is determinable automatically from syntax and morphological analysis, e.g., nominal case. And secondly, CATiB uses linguistic representation and terminology inspired by the long tradition of Arabic syntactic studies. This makes it easier to train annotators and not be restricted to hire annotators who have degrees in linguistics. This paper describes CATiB’s representation and compares it to other Arabic treebanking efforts. 1.
This paper presents a high-performance broad-coverage supervised word sense disambiguation (WSD) system for English verbs that uses linguistically motivated features and a smoothed maximum entropy machine learning model. We describe three specific enhancements to our system’s treatment of linguistically motivated features which resulted in the best published results on SENSEVAL-2 verbs. We then present the results of training our system on OntoNotes data, both the SemEval-2007 task and additional data. OntoNotes data is designed to provide clear sense distinctions, based on using explicit syntactic and semantic criteria to group WordNet senses, with sufficient examples to constitute high quality, broad coverage training data. Using similar syntactic and semantic features for WSD, we achieve performance comparable to that of human taggers, and competitive with the top results for the SemEval-2007 task. Empirical analysis of our results suggests that clarifying sense boundaries and/or increasing the number of training instances for certain verbs could further improve system performance.
We introduce a new type of lexical structure called lexical system, an interoperable model that can feed both monolingual and multilingual language resources. We begin with a formal characterization of lexical systems as simple directed graphs, solely made up of nodes corresponding to lexical entities and links. To illustrate our approach, we present data borrowed from a lexical system that has been generated from the French DiCo database. We later explain how the compilation of the original dictionary-like database into a net-like one has been made possible. Finally, we discuss the potential of the proposed lexical structure for designing multilingual lexical resources. [ABSTRACT FROM AUTHOR], Copyright of Language Resources & Evaluation is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual )
Chinese chunking plays an important role in natural language processing. This paper presents a large margin method for Chinese chunking based on structural SVMs (support vector machines). First, a sequence labeling model and the formulation of the learning problem are introduced for Chinese chunking problem, and then the cutting plane algorithm is applied to efficiently approximate the optimal solution of the optimization problem. Finally, an improved F1 loss function is proposed to tackle Chinese chunking. The loss function can scale the F1 loss value to the length of the sentence to adjust the margin accordingly, leading to more effective constraint inequalities. Experiments are conducted on UPENN Chinese Treebank-4 (CTB4), and the hamming loss function is compared with the improved F1 loss function. The experimental results show that the training algorithm with the improved F1 loss function can achieve higher performance than the Hamming loss function. The overall F1 score of Chinese chunking obtained with this approach is 91.61%, which is higher than the performance produced by the
The reporting of measures of effect size has become increasingly important in psychology. A Monte Carlo resampling permutation procedure is introduced to find near-optimum maximum values for Stuart’s τc measure for two-way ordinal contingency tables, also termed Kendall’s τc since Kendall introduced τa and τb . Comparisons between resampling and exact procedures demonstrate the accuracy and utility of resampling measures of effect size for two-way ordinal contingency tables. The resampling procedure is shown to be more precise than the traditional method of standardizing τc .
Computer-based diagnostic assessment systems hold potential to help teachers identify sources of poor performance and to connect teachers and students to learning activities designed to help advance students’ conceptual understandings. The present article presents findings from a study that examined how students’ performance in algebra and their overcoming of common algebraic misconceptions were affected by the use of a diagnostic assessment system that focused on important algebra concepts. This study used a four-group randomized cluster trial design in which teachers were assigned randomly to one of four groups: a “business as usual” control group, a partial intervention group that was provided with access to diagnostic tests results, a partial intervention group that was provided with access to the learning activities, and a full intervention group that was given access to the test results and learning activities. Data were collected from 905 students (6th–12th grade) nested within 44 teachers. We used hierarchical linear modeling techniques to compare the effects of full, partial, and no (control) intervention on students’ algebraic ability and misconceptions. The analyses indicate that full intervention had a net positive effect on ability and misconception measures.
The lexical development system OntoNet is introduced, which includes a browser and an editor for the WordNet 3.0 database. The aim of the OntoNet project is to provide a comfortable and up-to-date access to the lexical database for the modification of WordNet or the development of new wordnets.
This paper presents a novel approach of incorporating fine-grained treebanking decisions made by human annotators as discriminative features for automatic parse disambiguation. To our best knowledge, this is the first work that exploits treebanking decisions for this task. The advantage of this approach is that use of human judgements is made. The paper presents comparative analyses of the performance of discriminative models built using treebanking decisions and state-of-the-art features. We also highlight how differently these features scale when these models are tested on out-of-domain data. We show that, features extracted using treebanking decisions are more efficient, informative and robust compared to traditional features.
Abstract This article examines three critical issues relating to the role of culture in teaching English as an international language (EIL): Firstly, the way in which top-down processes of globalisation, accompanied by the widespread desire for English in many former colonial countries, have in general fostered the negative effects of dominance, divisiveness and difference in world social relations, resulting in the suppression and devaluation of local forms of knowledge and practice. Secondly, the way in which shifts of ownership and authority to non-native speakers and their varieties of English in combination more recently with global cultural flows, have created the need for reconceiving English as a pluralised global language, informed by local norms, functions and practices, reflecting a fluid and multiple cultural base. Lastly, it explores the way in which an ecological approach to English language teaching, which is oriented to ‘globalisation from below’ (Appadurai, 2000; Canagarajah, 2005), and which opens up a dialogical relationship between the global and the local, might help speakers in ex-colonial settings to reclaim their local identity and voice and thus realise the potential of globalisation to construct more inclusive, democratic relationships. Cet article examine trois problèmes cruciaux relatifs au rôle de la culture dans l'apprentissage de l'anglais en tant que langue internationale. Premièrement, il examine comment le processus ‘top-down’ de mondialisation, accompagné du désir répandu pour l'anglais dans plusieurs anciennes colonies a, en general, encouragé les effets négatifs de dominance, de dissension et de différence dans les rapports sociaux dans le monde. Le résultat en est la suppression et la dévalorisation des formes et des pratiques locales. Deuxièmement, il examine comment les déplacements de propriété et d'autorité qui désormais incombent les interlocuteurs non-natifs et leurs variétés d'anglais, depuis peu en cumul avec le flux culturel mondial, ont crée le besoin de reconceptualiser l'anglais comme une langue mondiale plurale alimentée en des normes, fonctions et pratiques locales reflétant une base culturelle multiple et fluide. Dernièrement, il explore comment une approche écologique à l'apprentissage de la langue anglaise, orientée vers la mondialisation par en-dessous (Appadurai, 2000; Canagarajah, 2005), et qui ouvre un rapport dialogique entre le mondial et le local, pourrait aider les interlocuteurs qui se trouvent dans les situations ex-coloniaux, à récuperer leurs identité et voix locales et ainsi à réaliser le potentiel de mondialisation afin d’établlir des rapports plus inclusifs et démocratiques. Keywords: English as an international language (EIL)globalisationlocal forms of knowledge and practiceculturehybridity Acknowledgements The author thanks Nirupama Rastogi for her translation of the abstract. Notes 1. I am aware that Phillipson's (1992) proposition that the spread of English is the result of a deliberate imperialist policy has been challenged by scholars who disagree with this argument. In Brutt-Griffler's (Citation2002), p. ix) view, English ‘owes its existence as a world language in large part to the struggle against imperialism, and not to imperialism alone’. Fishman (Citation1996, p. 640) maintains that the current spread of English is more linked to countries’ engagement in the modern world economy ‘than to any efforts derived from their former colonial masters’. Bhatia and Ritchie (Citation2004, p. 513) even hypothesise that, ‘The economic forces of globalization together with the rise of global media have set the stage for a dramatic, exponential rise in global bilingualism’. While arguments, both for and against linguistic imperialism, have their measure of validity, my concern here is far more with the (negative) consequences of the spread of English than its causes, and the conditions for redressing them. 2. Standards of English have remained a continuing concern at the highest levels of decision making in Singapore with Lee Kwan Yew, the former Prime Minister, and currently Minister Mentor, taking a keen, personal interest in these matters. 3. While it is true that the banning of local languages and literatures was part of the imperialist dominance practice of subjugation, Macaulay's (1952) now infamous Minute of the passing of the English Education Act in 1835 in colonial India, with which the then official languages, Persian and Arabic, were replaced by English, was brazenly justified on the ground that ‘a single shelf of European library is worth the whole native literature of India and Arabia’. Macaulay (Citation1952) declared that the native languages were neither moral nor scientific and thus were useless, much in the same way in which ‘foreign’ minority cultures and their ‘lack of civilisation’ were negatively constructed to justify colonisation. 4. Ngugi's (1986) recounting of one of the most humiliating experiences in a colonial school system where children caught speaking the local language were given the cane and forced to wear a placard round their necks with the inscription that read ‘I am a donkey’ or ‘I am stupid’ is painfully reminiscent of such elitism. 5. That is, although when used by ELF speakers English becomes a culture-free code (i.e. free of native-speaker culture), equally ELF is flexible enough to allow speakers to express their individual cultural identity – or ‘original voice’ (Kramsch, Citation1999) – by ‘exporting’ their L1 into the ELF and at the same time stimulate them to co-create a new inter-culture together with their co-participants. When Polzl notes that native-speaker English is ‘voice resistant’ he is referring to its intolerance of these dynamic aspects of the cultural, spatial and/or historical independence (no common ethnic origin) that characterise such interculture(s). 6. Scholars such as Kachru (Citation1992), Moag (Citation1992) and Schneider (Citation2003,Citation2007) agree in general terms that the new varieties of English pass through a series of evolutionary stages, starting with a reliance on exonormative varieties and a prejudice against the local variety through to the stage where the local variety receives local acceptance and becomes the classroom model. Schneider (Citation2003,Citation2007) identifies five phases: (1) the foundation phase when English first arrives; (2) the phase of exonormative stabilisation, characterised by the variety spoken by the settlers; (3) the nativisation phase when the variety of English takes on local lexical and cultural features; (4) the phase when the newly formed indigenous variety becomes accepted as the (endonormative) standard and (5) the final phase of differentiation signalling the emergence of different local varieties that mark ethnicity, identity and so on.
For Arabic, diacritizing written text is important for many NLP tasks. In the work presented here, we investigate the quality of a diacritization approach, with a high success rate for treebank data but with a more limited success on realworld data. One of the problems we encountered is the non-standard use of the hamza diacritic, which leads to a decrease in diacritization accuracy. If an automatic hamza restoration module precedes diacritization, the results improve from a word error rate of 9.20% to 7.38% in treebank data, and from 7.96% to 5.93% on selected real-world texts. This shows clearly that hamza restoration is a necessary step for improving diacritization quality for Arabic real-world texts.
Due to the lack of descriptive mechanism for syntactic functional structure,the dependency grammar can not express all complex syntactic structures explicitly.In addition,few parsing model takes the restrictive information of modifiers' nesting level into account,though it is a common sense in pragmatics.To solve these problems,a generative binary combinational grammar (BCG) parsing model is proposed which incorporates the restrictive information.In this model,the construction of sentence is regarded as the combination of adjacent chunks according to their headwords.Moreover,the symbolic local priorities between the adjacent binary relations and the modifiers' nesting levels are used to constrain the generation of parsing trees.The BCG parsing model is constructed by converting the dependency treebank to the BCG form.Then,the syntactic relations,the local priorities and the parameters of the model are induced automatically.Experimental results show that the proposed model improves the parsing accuracy.
Within the STEVIN1 project Large Scale Syntactic Annotation of written Dutch (LASSY), a manually corrected treebank of 1 million words is constructed. Lassy is part of a series of annotation projects for modern written and spoken Dutch. More specifically, it is an extension of the D-Coi and CGN projects,2 and constitutes the core of SoNaR, a 500 million words reference corpus of modern written Dutch.3 One of the goals of the latter project is to enrich the corrected treebank produced in Lassy4 with several semantic layers. For a general overview of the relations between D-Coi, Lassy and SoNaR, cf [19]. In this paper we will concentrate on the semantic layers of SoNaR core: (1) named entity labeling, (2) annotation of co-reference relations, (3) semantic role labeling and (4) annotation of spatial and temporal relations. Of these (2) originates from the STEVIN-project COREA,5 (3) and (4) from D-Coi, whereas (1) is a new area within STEVIN.
The Old Czech hapax legomenon bnedovánie is used in the adaptation of the medieval knight romance Duke Ernest (Herzog Ernst). There are two hypotheses based on the lexical system: 1) the corrupt word is an archaic part of the word family constituted by the adjectiv sbedný and its derivates and it means,a (brilliant and amusing) social behaviour corresponding with norms of court; (but we cannot satisfactorily explain word-formative processes leading from the sbedný to the bnedovánie); 2) the word is the corrupt form of Old Czech noun burdovánie (but its use is not convincing in the existing context).
This paper reports a sociolinguistic study of the state of Greek language in Australia as spoken by native-speaking Greek immigrants and their children. Emphasis is given to the analysis of the linguistic behaviour of these Greek Australians which are attributed to contact with English and to other environmental, social and linguistic influences. The paper discusses the non-standard phenomena in various types of inter-lingual transferences in terms of their incidence and causes and, in correlation with social, linguistic and psychological factors in order to determine the extent of language assimilation, attrition, and the content and context and medium of the language-event. The paper also discusses the transferences from English to Greek and vice- versa from a qualitative and quantitative perspective, of the phonemic, lexical, morphological, syntactic, semantic, pragmatic and prosodic deviations. During the last 170 years of settlement, Greek Australians know and use a new communicative norm with some degree of stability, the Ethnolect, (a non-standard variety of language used by an ethnic group in a static or dynamic bilingual situation) which serves their linguistic needs.
In the Computational Linguistics community, much work is put into the creation of large, high-quality linguistic resources, often with complex annotation. In order to make these resources accessible to non-technical audiences, formalisms for searching and filtering are needed, like the TIGER corpus query language. Recently, augmented treebanks have been published, including the SALSA corpus which features frame semantic annotation on top of syntactic structure. We design an extension for the TIGER language which allows searching for frame structures along with syntactic annotation. To achieve this, the TIGER ob ject model is expanded to include frame semantics, while remaining fully backwards-compatible and add these extensions to our own implementation of TIGER.
We describe the Hindi Discourse Relation Bank project, aimed at developing a large corpus annotated with discourse relations. We adopt the lexically grounded approach of the Penn Discourse Treebank, and describe our classification of Hindi discourse connectives, our modifications to the sense classification of discourse relations, and some crosslinguistic comparisons based on some initial annotations carried out so far.
The thesis studies the translation process for the laws of Finland as they are translated from Finnish into Swedish. The focus is on revision practices, norms and workplace procedures. The translation process studied covers three institutions and four revisions. In three separate studies the translation process is analyzed from the perspective of the translations, the institutions and the actors. The general theoretical framework is Descriptive Translation Studies. For the analysis of revisions made in versions of the Swedish translation of Finnish laws, a model is developed covering five grammatical categories (textual revisions, syntactic revisions, lexical revisions, morphological revisions and content revisions) and four norms (legal adequacy, correct translation, correct language and readability). A separate questionnaire-based study was carried out with translators and revisers at the three institutions. \n\nThe results show that the number of revisions does not decrease during the translation process, and no division of labour can be seen at the different stages. This is somewhat surprising if the revision process is regarded as one of quality control. Instead, all revisers make revisions on every level of the text. Further, the revisions do not necessarily imply errors in the translations but are often the result of revisers following different norms for legal translation. \n\nThe informal structure of the institutions and its impact on communication, visibility and workplace practices was studied from the perspective of organization theory. The results show weaknesses in the communicative situation, which affect the co-operation both between institutions and individuals. Individual attitudes towards norms and their relative authority also vary, in the sense that revisers largely prioritize legal adequacy whereas translators give linguistic norms a higher value. Further, multi-professional teamwork in the institutions studied shows a kind of teamwork based on individuals and institutions doing specific tasks with only little contact with others. This shows that the established definitions of teamwork, with people co-working in close contact with each other, cannot directly be applied to the workplace procedures in the translation process studied. Three new concepts are introduced: flerstegsrevidering (multi-stage revision), revideringskedja (revision chain) and normsyn (norm attitude). \n\nThe study seeks to make a contribution to our knowledge of legal translation, translation processes, institutional translation, revision practices and translation norms for legal translation. \n\n Keywords: legal translation, translation of laws, institutional translation, revision, revision practices, norms, teamwork, organizational informal structure, translation process, translation sociology, multilingual.
The process of marking up the syntactic structure of the sentences in a corpus is facilitated by having a graphical tree editor. A type specification constraints the nature of acceptable trees, and visual indication of non-conformities makes corrections easy. Projectivity checking can also be a useful way to notice errors, depending on the linguistic design of the tree type specification. A small test corpus has been marked up using this method. Only the English language has been investigated so far, and the tool takes no other input than the type specification and the corpus. Adding a lexical database is an obvious way to make it more generally useful.
Colloquial speech as any other sublanguage is subject to its own norm. Specificity of colloquial norm lies in its implicit character. Colloquial lexical units, which are normative for colloquial speech, are marked by minimal degree of stylistic degradation. In the process of unofficial communication the choice of linguistic units is subconsciously made within certain normative criteria. It is realized in case of violation of colloquial norm.
This paper presents a system for querying treebanks. The system consists of a powerful query language with natural support for cross-layer queries, a client interface with a graphical query builder and visualizer of the results, a command-line client interface, and two substitutable query engines: a very efficient engine using a relational database (suitable for large static data), and a slower, but paralel-computing enabled, engine operating on treebank files (suitable for "live" data).
We present a cost effective strategy for the creation of a mid-size fine-grained dependency treebank of surface- and deep-syntactic structures as defined in the Meaning-Text Theory for Spanish. The strategy starts from a small seed dependency corpus, the AnCora corpus, whose annotation is considerably more coarse-grained than our target annotation. We show that this discrepancy can be bridged largely by automatic means, relying upon contextual information and leaving thus minimal work to the annotators. This allows us to develop the resources with limited human effort within a limited period of time. We also propose a preliminary evaluation of the actual amount of work that the annotation process requires. 1
We describe for dependency parsing an annotation adaptation strategy, which can automatically transfer the knowledge from a source corpus with a different annotation standard to the desired target parser, with the supervision by a target corpus annotated in the desired standard. Furthermore, instead of a hand-annotated one, a projected treebank derived from a bilingual corpus is used as the source corpus. This benefits the resource-scarce languages which haven't different handannotated treebanks. Experiments show that the target parser gains significant improvement over the baseline parser trained on the target corpus only, when the target corpus is smaller.
Alternating Structure Optimization (ASO) is a recently proposed linear Multitask Learning algorithm. Although its effective has been verified in both semi-supervised as well as supervised methods, yet they necessitate taking external resource as a prerequisite. Therefore, feasibility of employing ASO to further improve the performance merely rests on the labeled data on hand proves to be a task deserving close scrutiny. Catering to this challenging while untapped problem, this paper presents a novel application of ASO to the subtask of Shallow Semantic Parsing: Chunking. Our experiments on Chinese Treebank 5.0 present promising result in chunk analysis, and the error rate is reduced by 5.72%, proposing a profound way to further improve the performance.
In this paper we describe and evaluate a top-down transfer component of a hybrid example-based machine translation system with an architecture similar to that of transfer MT systems, but with automatically derived transfer-rules and dictionary entries based on a parallel treebank. The tests were applied on the translation pair Dutch to English. Evaluation and error analysis have shown that the top-down transfer process has a number of shortcomings on which we wish to report and which we will try to solve in future work by applying bottom-up transfer.
We present a valency lexicon for Latin verbs extracted from the Index Thomisticus Treebank, a syntactically annotated corpus of Medieval Latin texts by Thomas Aquinas.
Written text is one of the fundamental manifestations of human language, and the study of its universal regularities can give clues about how our brains process information and how we, as a society, organize and share it. Among these regularities, only Zipf's law has been explored in depth. Other basic properties, such as the existence of bursts of rare words in specific documents, have only been studied independently of each other and mainly by descriptive models. As a consequence, there is a lack of understanding of linguistic processes as complex emergent phenomena. Beyond Zipf's law for word frequencies, here we focus on burstiness, Heaps' law describing the sublinear growth of vocabulary size with the length of a document, and the topicality of document collections, which encode correlations within and across documents absent in random null models. We introduce and validate a generative model that explains the simultaneous emergence of all these patterns from simple rules. As a r)
Increasing the domain of locality by using tree-adjoining-grammars (TAG) encourages some researchers to use it as a modeling formalism in their language application. But parsing with a rich grammar like TAG faces two main obstacles: low parsing speed and a lot of ambiguous syntactical parses. We uses an idea of the shallow parsing based on a statistical approach in TAG formalism, named supertagging, which enhanced the standard POS tags in order to employ the syntactical information about the sentence. In this paper, an error-driven method in order to approaching a full parse from the partial parses based on TAG formalism is presented. These partial parses are basically resulted from supertagger which is followed by a simple heuristic based light parser named light weight dependency analyzer (LDA). Like other error driven methods, the process of generation the deep parses can be divided into two different phases: error detection and error correction, which in each phase, different completion heuristics applied on the partial parses. The experiments on Penn Treebank show considerable improvements in the parsing time and disambiguation process.
Language can serve as a potent and injurious tool against ambitious women in the workplace. Consider how often both men and women have thought or said 'bitch' or other derogatory terms when describing a woman in a leadership position. Freud once described mental health as the ability to love and to work. For many of these women, their work has also become their love, their passion. Working long hours, often working 'twice as hard as men,' to be viewed as equally competent, perhaps sacrificing spouse and children in deference to career, these women have been inculcated into a patriarchal corporate environment, one in which they must act in a sexual dissonant manner in order to succeed. And as they negotiate the male managerial model, they often come to view their own femininity in an objectified, disparaging way. They may defeminize their language and adopt a 'more adversarial, information-focused style characteristic of all male talk'. Here is where the dilemma begins: If they conform to the masculine linguistic norms of the corporate environment, then their behavior can be perceived as confrontational, harsh, contentious, un-lady-like. A collision of culture ensues on intrapsychic and interpersonal levels between what is expected of a woman in society and what is expected of a person in a high status position where valued leadership traits are male-gendered. Often these women select language and styles that veer away from the feminine side of the gender spectrum. (PsycINFO Database Record (c) 2019 APA, all rights reserved)
Max Planck Institute for Psycholinguistics, Nijmegen, The Netherlands We present a coding system combined with an annotation tool for the analysis of gestural behavior. The NEUROGES coding system consists of three modules that progress from gesture kinetics to gesture function. Grounded on empirical neuropsychological and psychological studies, the theoretical assumption behind NEUROGES is that its main kinetic and functional movement categories are differentially associated with specific cognitive, emotional, and interactive functions. ELAN is a free, multimodal annotation tool for digital audio and video media. It supports multileveled transcription and complies with such standards as XML and Unicode. ELAN allows gesture categories to be stored with associated vocabularies that are reusable by means of template files. The combination of the NEUROGES coding system and the annotation tool ELAN creates an effective tool for empirical research on gestural behavior.
OXlearn is a free, platform-independent MATLAB toolbox in which standard connectionist neural network models can be set up, run, and analyzed by means of a user-friendly graphical interface. Due to its seamless integration with the MATLAB programming environment, the inner workings of the simulation tool can be easily inspected and/or extended using native MATLAB commands or components. This combination of usability, transparency, and extendability makes OXlearn an efficient tool for the implementation of basic research projects or the prototyping of more complex research endeavors, as well as for teaching. Both the MATLAB toolbox and a compiled version that does not require access to MATLAB can be downloaded from http://psych.brookes.ac.uk/oxlearn/.
Atypical prosody has been identified as a core feature of Autism Spectrum Disorders (ASD). Even when other aspects of language improve, prosodic deficits tend to be persistent. Deficits in prosody may limit the social acceptance of children with High-Functioning Autism (HFA) mainstreamed into the larger community.Prosody in ASD is an underresearched and criticized area in general and there has been little research on the prosody of Israeli Hebrew (IH) and even fewer studies comparing the prosody of typical and atypical Hebrewspeaking children in particular.Our study compares and contrasts the intonation units (IU), simple pitch accents (PA), and edge tones (ET) of five children between 9 and 12 years of age diagnosed with HFA and five children without developmental disorders (WDD) in reading aloud and spontaneous speech elicitation tasks. The subjects were matched for age, year of school, and academic achievements and all were male monolingual speakers of IH.The data were transcribed using the Autosegmental-Metrical (AM) theory of intonation with the IH ToBI (Tones and Break Indices) system being developed for this study with the computerized PRAAT system. The results were analyzed and explained according to: (1) the defintion that language is a symbolic tool whose structure is shaped both by its communication function and by the characteristics of its users and (2) the principle that language represents a compromise in the struggle to achieve maximum communication through minimal effort associated with the theory of Phonology as Human Behavior (PHB).The children with HFA produced more IU and PA than the WDD children. The HFA children acquired a limited repertoire of prosodic-edgetone patterns within the norm of the language. These patterns were repeatedly used both in spontaneous speech and in the reading tasks. In contrast, the WWD control group used a greater number of prosodic patterns showing a larger degree of variation for the same speech and language tasks. This study has become the basis for further ongoing research which has shown clear parallels in the extralinguistic, paralinguistic (prosody), and linguistic (lexical repetition) behavior of HFA children.
The referential expressions of online sales clothing generally consist of the central word and multiple attributives.Central words are always created in recent years.Multiple attributives are rich in content and varied in word order.Referential expressions of online sales clothing reflect some cultural trends in our society.However,some of the expressions deviate from linguistic norms and call for standardization.
Text Resources and Lexical Knowledge is comprised of a selection of 19 papers presented at the 9th biennial conference on Natural Language Processing held in September–October 2008 at the Berlin-Brandenburg Academy of Sciences and Humanities in Berlin. The 19 papers are grouped in 3 different sections: ‘Linguistic Analysis of Text Resources’, ‘Extraction of Lexical Knowledge from Text Resources’, and ‘Representation of Lexical Knowledge’. Several of these papers will interest readers of the International Journal of Lexicography. The first section includes 6 papers. Marek et al.'s paper on Extending the TIGER query language with universal quantification shows how the TIGERSearch query language has been integrated in a parallel treebank exploration tool, which is used for cross-language comparisons and translation studies. The emphasis is on complex universal quantification, so that users can formulate queries such as ‘Find sentences that do not include word X’. The second chapter is a nice and clear paper by Schwarz, Bartsch, Eckart and Teich on Exploring automatic theme identification: a rule-based approach. The rule-based system presented by the authors makes it possible to automatically identify the theme of a sentence, defined as the point of departure of the clause as message, for corpus annotation, using a pattern-matching algorithm.
Abstract Recently, there has been a growing interest in regional variation within African American English. This study reviews a work done on local speech in Pittsburgh, Pennsylvania, discussing trends for both African American and White ethnic groups. Just as scholars have found in other geographic regions, in Pittsburgh, African Americans and Whites share a number of feature characteristics of the local dialect, but remain distinct in a number of other ways. Research in Pittsburgh, as elsewhere, highlights the complexity, rather than the homogeneity, of African American speech across the country, as speakers exhibit alignment to both regional and supraregional ethnic linguistic norms.
Bond ratings on state‐issued debt provide a signal to credit markets that help them charge an appropriate interest rate, based on the risk of payment default. Though actual default may occur only in extreme circumstances, observed differences in ratings and interest costs across states and time demonstrate that a sound economy, strong financials, and stable policies matter. When data on the factors that presumably affect ratings is public and easily accessible, making sense of differences of opinion between bond rating agencies is difficult. We suggest that such differences—observed as so‐called split bond ratings—are often ephemeral. Utilizing a simulation method to uncover the latent credit risk presented by each state, we show that split ratings on state bonds are often due to the fact that presumed category overlap between rating agencies is absent when evaluated on a common latent scale. Most observed state bond rating splits from 1997 through 2006 can be explained by this category mismatch. Our approach has broad implications for pricing state debt, as well as pricing rated debt in other capital market sectors.
A known problem of WordNet is that it is too ne-grained in its sense denitions. For instance, it does not distinguish between homographs and polysemes. This distinction is crucial in many natural language processing tasks. In this paper we propose to distinguish only between homographs withinWordNet data while merging all polysemous senses. The ultimate goal of this exercise is to compute a more coarsegrained version of linguistic database. In order to achieve this task we propose to merge all polysemous senses according to similarity scores computed by a hybrid algorithm. The key idea of the algorithm is to combine the similarity scores produced by diverse semantic similarity algorithms. We implemented the algorithm and evaluated it on the dataset extracted from the WordNet. The evaluation results are promising in comparison to the other state of the art approaches.