53,057 research outputs found
Arabic learner corpus and its potential role in teaching Arabic to non-native speakers
No full textThe literature on learner corpora (Al-Sulaiti, 2010; Granger & Dumont, 2012) shows that there is a need to compile an Arabic learner corpus, which can be used in research on Arabic language learning and teaching. This paper introduces the project to build the Arabic Learner Corpus (ALC) (Alfaifi & Atwell, 2013), and its potential uses in teaching Arabic as a second language (ASL). ALC includes 282,732 words of written and spoken materials produced by learners of Arabic in Saudi Arabia. The corpus covers two types of students, non-native Arabic speakers (NNAS) and native Arabic speaking students (NAS) who are learning to improve their written Arabic. The corpus also covers two general levels of study, pre-university and university. The corpus data is available online for download in TXT and XML format at http://www.arabiclearnercorpus.com . Further work is being performed to annotate the entire corpus for errors. The growing role of learner corpora in the last decade has provided valuable insights into the uses of this type of corpora, such as error analysis, contrastive interlanguage analysis, making learners’ dictionaries, monitoring learners’ improvements and designing language materials. These uses were considered when building ALC as an open-source corpus
A Word Hypothesis Lattice Corpus - a benchmark for linguistic constraint models
No full textIntroduction In response to an input sentence, a typical recognition system (be it speech or handwriting) will build and output a word-hypothesis recognition lattice. That is a sequence of word candidate sets, where each word position is composed of a number of candidate words proposed by the recogniser. For example, on `hearing' the sentence "Stephen left school last year", an English speech recognition system might produce the following lattice: Stephen stiffens left lift loft school scowl scull lest last lust least yearn your year To disambiguate such a lattice, a standard technique is to use a language model to constrain the possible choices, so that the chosen sequence of words is the most linguistically plausible (See, for example, Jelinek 90, Atwell et al 93, Rose and Evitt 92, Keenan 93, also other papers in this Proceedings). Unfortunately, no standard method exists for evaluating and comparing the relative effectiveness of the various linguistic con
Increasing our ignorance of language: identifying language structure in an unknown signal
No full textThis paper describes algorithms and software developed to characterise and detect generic intelligent language-like features in an input signal, using natural language learning techniques: looking for characteristic statistical "language-signatures" in test corpora. As a first step towards such species-independent language-detection, we present a suite of programs to analyse digital representations of a range of data, and use the results to extrapolate whether or not there are language-like structures which distinguish this data from other sources, such as music, images, and white noise. Outside our own immediate NLP sphere, generic communication techniques are of particular interest in the astronautical community, where two sessions are dedicated to SETI at their annual International conference with topics ranging from detecting ET technology to the ethics and logistics of message construction (Elliott and Atwell, 1999; Ollongren, 2000; Vakoch, 2000)
Traumatic identity and aura in David Lodge's "Author, author"
No full textEste artículo analiza la novela Author, Author (2004) de David Lodge como ejemplo de bioficción neo-victoriana centrada en una celebridad, en este caso concreto, Henry James. El género forma parte del renacimiento Victoriano actual que afecta a los estudios culturales en su conjunto. Mi argumento central es que la novela de Lodge constituye una respuesta a las ansiedades culturales actuales, en particular a las que se refieren a la crisis identitaria y autoría literaria, así como a la pérdida del aura artística de Walter Benjamin, sublimándolas a través de los traumas de finales del siglo XIX. La elección de James, como demuestra el artículo, no es casual. Es el último representante de un mundo perdido en el que el aura aún tenía un espacio; el ser humano en crisis y traumatizado porque no encaja en un status quo nuevo.This paper delves into David Lodge’s Author, Author (2004) as an example of neo-Victorian celebrity biofiction, more concretely on Henry James. The genre belongs to the wave of Victorian revival in current literature which also affects cultural studies in general. My main contention is that Lodge’s novel responds to current cultural anxieties, particularly the crisis of identity and authorship and the end of Walter Benjamin’s concept of aura, by sublimating them into late-nineteenth-century traumata. The choice of James is, the article argues, not casual. He represents the redeeming figure of a lost auratic world; the human in crisis, traumatized because he does not fit in the new status quo
A generic template to evaluate integrated components in spoken dialogue systems
No full textWe present a generic template for spoken dialogue systems integrating speech recognition and synthesis with 'higher-level' natural language dialogue modelling components. The generic model is abstracted from a number of real application systems targetted at very different domains. Our research aim in developing this generic template is to investigate a new approach to the evaluation of Dialogue Management Systems. Rather than attempting to measure accuracy/speed of output, we propose principles for the evaluation of the underlying theoretical linguistic model of Dialogue Management in a given system, in terms of how well it fits our generic template for Dialogue Management Systems. This is a measure of 'genericness' or 'application-independence' of a given system, which can be used to moderate accuracy/speed scores in comparisons of very unlike DMSs serving different domains. This relates to (but is orthogonal to) Dialogue Management Systems evaluation in terms of naturalness and like measurable metrics (eg Dybkjaer et al 1995, Vilnat 1996, EAGLES 1994, Fraser 1995); it follows more closely emerging qualitative evaluation techniques for NL grammatical parsing schemes (Leech et al 1996, Atwell 1996)
Detecting student copying in a corpus of science laboratory reports
No full textThis case study is an evaluation of generic, general-purpose plagiarism detection systems applied to a specific domain and task: detecting intra-class student copying in a corpus of Biomedical Science laboratory reports. From the outset, our project had the practical, pragmatic aim to find a workable solution to a specific problem. Biomedical Science undergraduates learn experimental methods by working through a series of laboratory experiments and reporting on their results. These laboratory reports are “peer-reviewed” in large classes, following a prescribed marking scheme; as the reports are effectively marked by other students rather than by a single lecturer, there is an opportunity for an unscrupulous student to avoid having to carry out and report on an experiment, by simply copying another student’s report. To reduce this temptation, the Biomedical Science director of teaching, Paul Gent, approached Eric Atwell of the School of Computing and Clive Souter of the Centre for Joint Honours in Science, to look at ways to compare laboratory reports automatically, and flag candidates with signs of copying. We were joined by Julia Medori, forensic linguist from Trinity College Dublin, who developed and evaluated a range of possible solutions
A lexical database for English learners and users: the Oxford advanced learner's dictionary
No full textA cross-language methodology for corpus part-of-speech tag-set development
No full textThis paper examines criteria used in development of Corpus Part-of-Speech tag sets used when PoS-tagging a corpus, that is, enriching a corpus by adding a part-of-speech category label to each word. This requires a tag-set, a list of grammatical category labels; a tagging scheme, practical definitions of each tag or label, showing words and contexts where each tag applies; and a tagger, a program for assigning a tag to each word in the corpus, implementing the tag-set and tagging-scheme in a tag-assignment algorithm
Development of tag sets for part-of-speech tagging
No full textThis article discusses tag sets used when PoS-tagging a corpus, that is, enriching a corpus by adding a part-of-speech tag to each word. This requires a tag-set, a list of grammatical category labels; a tagging scheme, practical definitions of each tag or label, showing words and contexts where each tag applies; and a tagger, a program for assigning a tag to each word in the corpus, implementing the tag-set and tagging-scheme in a tag-assignment algorithm. We start by reviewing tag-sets developed for English corpora in section 1, since English was the first language studied by corpus linguists. Pioneering corpus linguists thought that their English corpora could be more useful research resources if each word was annotated with a Part-of-Speech label or tag. Traditional English grammars generally provide 8 basic parts of speech, derived from Latin grammar. However, most tag-set developers wanted to capture finer grammatical distinctions, leading to larger tag-sets. PoS-tagged English corpora have been used in a wide range of applications. Section 2 examines criteria used in development of English corpus Part-of-Speech tag sets: mnemonic tag names; underlying linguistic theory; classification by form or function; analysis of idiosyncratic words; categorization problems; tokenisation issues: defining what counts as a word; multi-word lexical items; target user and/or application; availability and/or adaptability of tagger software; adherence to standards; variations in genre, register, or type of language; and degree of delicacy of the tag-set. To illustrate these issues, section 3 outlines a range of examples of tag set developments for different languages, and discusses how these criteria apply. First we consider tag-sets for an online Part-of-Speech tagging service for English; then design of a tag-set for another language from the same broad Indo-European language family, Urdu; then for a non-Indo-European language with a highly inflexional grammar, Arabic; then for a contrasting non-Indo-European language with isolating grammar, Malay. Finally, we present some conclusions in section 4, and references in section 5
- …
