Dissertations / Theses on the topic 'Language sampling'

Many probabilistic inference algorithms, both exact and approximate, have been developed to run efficiently on various probabilistic models in the recent decades. For many probabilistic models, exact inference is, however, infeasible or impossible. As such, approximate algorithms are often necessary. In this thesis, a method for partially applying exact inference in probabilistic programming languages using Monte Carlo inference algorithms is introduced and formalized. More specifically, this method allows for conditioning on observations in the probabilistic program before performing sampling, where applicable. We show that this method, called delayed sampling, can be used to reduce mean squared error of estimators based on samples generated by probabilistic programs. We also show that delayed sampling never leads to an increase in mean squared error of estimators. An evaluation is performed with an implementation of delayed sampling in the probabilistic programming language Anglican. The results demonstrate clear reductions in mean squared error for simple examples, but the overhead is also shown to be quite substantial for the Anglican implementation.
Många probabilistiska inferensalgoritmer, både exakta och approximativa, har utvecklats för att fungera effektivt på olika probabilistiska modeller under de senaste decennierna. För många probabilistiska modeller är exakt inferens emellertid otänkbar eller omöjlig. På grund av detta är ofta approximativa algoritmer nödvändiga. I denna avhandling introduceras och formaliseras en metod för att delvis tillämpa exakt inferens i probabilistiska programmeringsspråk med Monte Carlo-inferensalgoritmer. Mer specifikt tillåter denna metod att konditionera på observationer i det probabilistiska programmet innan provtagning utförs, där så är tillämpligt. Vi visar att den här metoden, som kallas fördröjd provtagning, kan användas för att minska genomsnittliga kvadratiska fel för estimatorer som baseras på prover genererade av probabilistiska program. Vi visar också att fördröjd provtagning aldrig leder till en ökning av genomsnittliga kvadratiska fel för estimatorer. En utvärdering utförs med en implementering av fördröjd provtagning i det probabilistiska programmeringsspråket Anglican. Resultaten visar tydliga minskningar i genomsnittligt kvadratfel för enkla exempel, men beräkningskostnaderna visar sig också vara ganska betydande för implementationen i Anglican.

The current study examined two methods of language sampling (interview and the Dixit Method) in early adolescents with typical language development. The purpose of this study was to examine the effectiveness of the two procedures in eliciting lexical diversity in spoken language samples in the early adolescent population. To examine lexical diversity, traditional methods of analysis (type token ratio, average type token ratio, and mean length of utterance) were applied. However, because literature historically alludes to flaws in these traditional methods, the researchers also applied an ecological approach to analysis of lexical diversity proposed by Scott Jarvis (2013). Student participants (n = 22) in the sixth grade (11:0 -12:11 years of age) were recruited through a local middle school. Both methods of language sampling were compared in a within-subject design for diversity of spoken language samples by applying the previously mentioned analysis procedures. The data was evaluated using a pairedsamples two-tailed t test. Although the sample size was small, evidence from this study indicates the Dixit Method provides a more holistic view of lexical diversity than the traditional method by considering six ecological components that are reflected in typical language use.

Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2017.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 95-96).
Implementing Monte Carlo integration requires significant domain expertise. While simple algorithms, such as unidirectional path tracing, are relatively forgiving, more complex algorithms, such as bidirectional path tracing or Metropolis methods, are notoriously difficult to implement correctly. We propose a domain specific language for Monte Carlo rendering that offers primitives and data structures for writing concise and correct-by-construction sampling code. The compiler then automatically generates the necessary code for evaluating PDFs and combining multiple samples. Our language focuses on ease of implementation for rapid exploration and research, at the cost of run time performance. We demonstrate the effectiveness of the language by implementing several challenging rendering algorithms, as well as a new algorithm, which would otherwise be prohibitively difficult to implement.
by Luke Anderson.
S.M.

The present study sought to determine the effect different stimulus material has on the language elicited from children. Its purpose was to determine whether a significant difference existed among language samples elicited three different ways when analyzed using DSS. Eighteen children between the ages of 3.6 and 5.6 years were chosen to participate in the study. All of the children had normal bearing. normal receptive vocabulary skills and no demonstrated or suspected physical or social delays. Three language samples. each elicited by either toys. pictures. or stories. were obtained from each child. For each sample. a corpus of 50 utterances was selected for analysis and analyzed according to the DSS procedure as described by Lee and Ganter (1971).

The current study examined a method of language sampling (the Dixit Method- Science, Math, Engineering, Arts, and Math) in early adolescents with typically developing language. The purpose of this study was to examine the effectiveness of the DM-STEAM in eliciting lexically sophisticated spoken language samples in the early adolescent population. To examine lexical sophistication, traditional measures of analysis such as mean length of utterance (MLU) and average type token ratio (AVG TTR) were applied along with a measure of low frequency vocabulary. To compare performance on the DM-STEAM, school standardized assessments were obtained to measure student skill in academic content areas. Twenty-two student participants in the sixth grade (11 years to 12 years 11 months) were recruited from a local elementary school. The data was evaluated using a paired tailed t test and a path analysis test. Although the sample size is small, results from the study indicate the DM-STEAM elicits low frequency academic vocabulary in early adolescent populations.

Young children begin to draw spontaneously, and their drawings develop in complexity and skill as they become older. Previously, children's drawings have been utilized to assess aspects of emotional and cognitive development. On the other hand, very few studies have used children's drawings as a tool for understanding their language development. The purpose of this study was to examine the potential for using drawing as an assessment tool for language skill. Therefore, this study explored the relationship between pre-kindergarten children's drawing and language development and stability of this relationship over time. This study also examined whether children's English-language status influenced the drawing-language relationship. The study's subjects were 34 children who attended a pre-kindergarten program in a local elementary school located in a Mid-Western university town. These children participated in take-home reading and drawing activities once a week for one school year. Children's drawings that were completed after reading storybooks with their parents served as data documenting their drawing skills. Classroom teachers' observations of the children's language use were recorded on the Work Sampling System's developmental checklists, and this served as data to document their language skills. A hierarchical regression analysis revealed that children's drawing skill was a significant predictor to their language skill and explained 10% of the variance in language scores. This finding was consistent at both the beginning and end of the school year. Children's English-language status did not make a statistically significant difference in the predictive relationship of drawing for their language skills. Therefore, this study concluded that children's drawings can be used as a supplementary assessment tool to provide information about children's language skills, regardless of their primary language. However, more than half of the language variance was not identified from this study; this is an area for further investigation.

U  radu  je  analizirana  brioflora  Fruške  gore.  Cilj istraživanja bio je  utvrđivanje diverziteta, distribucije, uslova mikrostaništa i strukture zajednica mahovina na  različitim  tipovima  staništa  na  Fruškoj  gori. Takođe,  jedan  od  ciljeva  je  bio  i  testiranje  i optimizacija  metode  za  kvantitativna  uzorkovanja briofita  na  različitim  tipovima  staništa.  Terenska istraživanja su vršena u periodu 2013-2017 godine. Zapotrebe florističkih istraživanja terenska uzorkovanja su  vršena  metodom  transekta  na  120  lokaliteta  na području  Fruške  gore.  Za  testiranje  i  optimizaciju metoda  za  kvantitativna  briološka  uzorkovanja korišćene  su  različite  varijante  metoda  „kvadrata“ (metod ugneždenih kvadrata za prizemnu briofloru i metod  postavljanja  mikroplotova  na  različitoj udaljenosti  za  epifitsku  briofloru).  Odgovarajuće metode  su  odabrane  na  osnovu  kriva  minimum areala  i  kriva  sličnosti  (eng.  species-area  curve, similarity  area  curve).  Odgovarajuće  metode  su primenjene za dalja uzorkovanja u ovom istraživanju. Na pojedinim lokalitetima (šumska i livadska staništa) vršeno je merenje nekoliko ekoloških parametara (pH zemljišta  i  kore  drveta,  vlažnost  zemljišta, temperatura zemljišta, pokrovnost stelje, pokrovnostzeljastih vaskularnih biljaka, udaljenost od potoka u šumskim staništima i udaljenost šumskih ekosistema na  livadskim  staništima)  u  cilju  utvrđivanja  uticaja merenih  ekoloških  parametara  na  diverzitet  i distribuciju briofita.  Briofitske zajednice su utvrđene na  različitim  tipovima  staništa  i  podloge  uz  pomoć „label  propagating  community  detection“  analize.Rezultati  ovog  istraživanja  su  pokazali  da  je  Fruškagora područje koje se odlikuje visokim diverzitetom briofita  i  ukupno  je  zabeleženo  235  briofitskih taksona (na osnovu literaturnih podataka i terenskih istraživanja).  73  taksona  briofita  je  prvi  put zabeleženo  na  području  Fruške  gore.  Od  ukupnogbroja  taksona,  11%  je  pod  nekim  stepenom ugroženosti ili zakonske zaštite. Metoda mikrocenoze se  pokazala  kao  najadekvatniji  metod  za kvantitativna  uzorkovanja  prizemnih  briofita  (sa minimalnom adekvatnom veličinom kvadrata 0.5x0.5 m  u  šumskim  odnosno  1x1  m  na  livadskim staništima). Za epifitske briofite, kao najadekvatnija se  pokazale  metoda  postavljanja  mikroplotova dimenzija  10x10  cm  sa  sve  četiri  strane  stabla  na međusobnoj udaljenosti od 20 cm. Struktura staništa ima  značajno  veći  uticaj  na  diverzitet  prizemnih briofita  u  šumskim  staništima  u  odnosu  na karakteristike  zemljišta,  dok  je  na  livadskimstaništima zabeležen obrnut slučaj.  Najveći uticaj na diverzitet epifitskih briofita ima visina na stablu kao i tip forofite. Na istraživanom području je detektovano devet  zajednica  u  prizemnoj  brioflori  šumskih staništa;  šest  epifitskih  zajednica;  pet  zajednica  na livadskim  staništima;  četiri  zajednice  na  trulim deblima  i  panjevima;  10  zajednica  na  stenama  i kamenju;  četiri  zajednice  u  potocima.  Dobijeni rezultati  predstavljaju  građu  za  dalja  briološka istraživanja, kako floristička tako i ekološka, i dopunu flore Srbije.
In  this research, the bryophyte flora of Fruška gora Mountain was analyzed. The aim of this  research was to determine the diversity, distribution, microhabitat conditions  and  community  structure  of  bryophytes on different types of habitats on Fruška Gora. Also, one of the goals was testing and optimization of the method for  quantitative sampling  of bryophytes on different types of habitats. Field research was    carried out  in  the  period  2013-2017.  For  the  purposes  of floristic research, field sampling was carried out using the  transect  method  at  120  localities.  Different variants of the "quadrat" method (method of nested quadrats  for terrestrial bryophytes  and the method of  placing  microplots  at  different  distances  for  the epiphytic  byphytes)  were  used  for  testing  and optimization of methods for quantitative  bryophyte sampling. Appropriate methods were selected on the basis  species-area  curve  and  similarity-area  curve. Appropriate  methods  have  been  applied  to  further sampling in this study. On some localities (forest and meadow habitats) several  environmental parameters (pH  of  soil  and  bark  of  wood,  soil  moisture,  soil temperature,  roof  cover,  coverage  of  herbaceous vascular  plants,  distance  from  the  stream  in  forest habitats  and  distance  of  forest  ecosystems  in meadow  habitats)  were  measured  in  order  to determine their  impact  on  the  diversity  and distribution  of  bryophytes.  Bryophyte  communities were  identified  on  different  substrates  and  habitat types  using  "label  propagating  community detection"  analysis.  The  results  of  this  study  have shown that Fruška gora is an area characterized by a high  diversity  of  bryophytes  and  a  total  number  of 235 taxa  were recorded (based on literature data and field  research).  73  taxa  were  recorded  for  the  first time on this area. Of the total number of taxa, 11% is threatened  or  under  legislative  protection.  The microcenose  method  proved  to  be  the  most adequate  for  quantitative  sampling  of  terrestrial bryophytes (with a minimum quadrat size of  0.5x0.5 m in forests and 1x1 m in meadow habitats). For the epiphytic  bryophytes,  the  method  of  placing  the microplots (10x10 cm) on all four sides of the tree at a  distance  of  20  cm  is  most  suitable.  The  stand structure  has a significantly greater influence on the diversity of terrestrial  bryophytes  in  forest  habitats compared  to  soil   haracteristics,  while  in  the meadow  habitats  there  is  a  reverse  case.  The greatest  influence  on  the  diversity  of  epiphytic bryophytes has height on the tree as well as the type of  phorphyte.  In  the  investigated  area,  nine bryophyte  communities  were  detected  in  ground bryopyhte  flora  of  forest  habitats;  six  epiphytic communities; five communities in meadow habitats; four  communities  on  rotten  trunks  and  horns;  10 communities  on  the  rocks  and  stones;  four communities  in  streams.  The  obtained  results represent the  base  for  further  bryological  research, both floristic and ecological, and the addition  to the flora of Serbia.

U doktorskoj disertaciji evaluirani su rezultati kvantifikovanih koncentracionih nivoa lipofilnih perzistentnih i emergentnih organskih polutanata u uzorcima sedimenta kolektovanih sa deset reprezentativnih lokaliteta u srednjem Podunavlju. Na osnovu sprovedenih laboratorijskih i terenskih istraživanja definisan je prostorni trend jedinjenja na ispitivanom području. Primenom multivarijantnih tehnika eksperimentalni rezultati su uspešno modelovani statističkim metodama koje su izdiferencirale izvore kontaminacije za ukupan set ispitivanih polutanata. Prikazana je procena uticaja kontaminiranog sedimenta na akvatične organizme i humanu populaciju prema stepenu kancerogenosti jedinjenja. Predstavljeni su različiti scenariji izloženosti ingestijom i dermalnim kontaktom, u zavisnosti od vremena ekspozicije i izložene površine potencijalnih receptora, različitog uzrasta i pola. Po prvi put u istraživanom području, implementirana je ex-situ ravnotežna metodologija pasivnog uzorkovanja primenom sorpcionog medijuma od silikonske gume testirane pri različitim masenim odnosima polimer-uzorak sedimenta. Primenjenom metodom uspešno su dobijene slobodno rastvorene koncentracije lipofilnh kontaminanata u pornoj vodi sedimenta, kao prediktora za određivanje biodostupnosti jedinjenja.
The doctoral dissertation evaluated the results of quantified concentration levels of lipophilic persistent and emergent organic pollutants in sediment samples collected from ten representative localities in the central Danube region. Based on the conducted laboratory and field research, the spatial distribution of compounds in the studied area was defined. Using multivariate techniques, the experimental results were successfully modeled by statistical methods that differentiated the sources of contamination for the total set of tested pollutants. A risk assessment of contaminated sediment on aquatic organisms and the human population according to carcinogenicity of the tested compounds is presented in thesis. Different exposure scenarios of ingestion and dermal contact, depending on the time exposure and exposed surface skin of potential receptors, of different ages and sex, are presented. For the first time in the study area, an ex-situ equilibrium passive sampling methodology was implemented using silicone rubber sorption medium tested at different polymer-sample sediment mass ratios. Using the method, the freely dissolved concentrations of lipophilic contaminants in pore water were successfully obtained as predictors for determining the bioavailability of the compounds.

Oral narrative ability has received increasing attention over the past three decades, and the importance of children's oral narrative skills to academic achievement has been well established. Children with reading disabilities are known to demonstrate difficulties in the ability to produce and comprehend oral narratives (Roth & Spekman, 1986; Snyder & Downey, 1991). However, the nature of the relationship between oral narrative ability and reading comprehension performance in children with reading disability is not clear. The experiments reported in this thesis aim to address this issue. The following questions are asked: 1) Do deficits in oral narrative ability contribute to reading comprehension difficulties? and 2) What is the likely direction of the relationship between aspects of oral narrative ability and reading comprehension performance in children with reading disabilities? Fourteen children (aged between 6;4 and 7;8 at the initial assessment) with mixed reading disability (MRD: i.e., children who demonstrate both word recognition and listening comprehension deficits) participated in the study. Their oral narrative skills were compared to those of their chronological age-matched peers with typical development (TD) and their reading-age-matched peers with typical development (RMTD). The study consisted of three phases: 1) A longitudinal phase in which the children's oral narrative performance was assessed on three occasions over a two-year period; 2) An intervention phase (using a nonequivalent pretest-posttest control group design) in which 10 children participated in an oral narrative intervention program that focused on enhancing children's story structure knowledge; and 3) A follow-up assessment phase conducted eight months post-intervention. Oral narratives were elicited in a personal narrative context and in a story retelling context. Oral narrative production ability was analysed at macrostructure (story quality) and microstructure (morpho-syntax and vocabulary) levels. Oral narrative comprehension was assessed in a fictional story context through questions relating to story structure elements. Reading comprehension performance was assessed using a standardised test of reading ability. The results from the longitudinal study showed that the children with MRD demonstrated inferior oral narrative production and oral narrative comprehension performance compared to their peers with typical reading development at each assessment occasion. When comparing the poor readers' performance to the RMTD group at the third assessment trial, the results suggested that the children with MRD demonstrated a specific deficit in oral narrative comprehension. In contrast, a pattern of delay was observed on the microstructure measures of oral narrative performance. The results from the intervention indicated significant treatment effects with large effect sizes for oral narrative comprehension performance. Despite this improvement in oral narrative comprehension, there was little change in oral narrative production ability, and transfer to reading comprehension was not evident. Although the follow-up assessment indicated sustained improvement in oral narrative comprehension for the children with MRD, accelerated reading comprehension progress was not evident. The findings from the longitudinal case study highlighted the benefits of oral narrative intervention for a child considered at high risk of continuing academic difficulties. This thesis provides evidence of the persistent oral narrative difficulties in children with MRD. The findings also provide support for the importance of narrative structure knowledge to these poor readers' oral narrative comprehension performance. The results demonstrate that oral narrative comprehension ability explains only a small amount of the variance in reading comprehension performance. Rather, the persistent word recognition difficulties of the children with MRD exert the biggest influence on their reading comprehension performance. These results are discussed in terms of current models of reading and language development. Implications for clinical practice are also addressed.

Active learning reduces annotation costs for supervised learning by concentrating labelling efforts on the most informative data. Most active learning methods assume that the model structure is fixed in advance and focus upon improving parameters within that structure. However, this is not appropriate for natural language processing where the model structure and associated parameters are determined using labelled data. Applying traditional active learning methods to natural language processing can fail to produce expected reductions in annotation cost. We show that one of the reasons for this problem is that active learning can only select examples which are already covered by the model. In this thesis, we better tailor active learning to the need of natural language processing as follows. We formulate the Unreliable Parameter Principle: Active learning should explicitly and additionally address unreliably trained model parameters in order to optimally reduce classification error. In order to do so, we should target both missing events and infrequent events. We demonstrate the effectiveness of such an approach for a range of natural language processing tasks: prepositional phrase attachment, sequence labelling, and syntactic parsing. For prepositional phrase attachment, the explicit selection of unknown prepositions significantly improves coverage and classification performance for all examined active learning methods. For sequence labelling, we introduce a novel active learning method which explicitly targets unreliable parameters by selecting sentences with many unknown words and a large number of unobserved transition probabilities. For parsing, targeting unparseable sentences significantly improves coverage and f-measure in active learning.

Sprovedeno istraživanje u okvirudisertacije obuhvata procenu distribucije16 EPA primarnih policikličnih aromatičnihugljovodonika između gasovite i čvrstefaze u atmosferi, na selektovanimlokalitetima urbanog, industrijskog iruralnog područja na teritoriji ZapadnogBalkana. Rezultati eksperimentalnogodređivanja predstavljali su osnovu zaevaluaciju reprezentativnih modelarasprostiranja poluisparljivih organskihjedinjenja kroz površinski atmosferski sloji raspodele gas/suspendovane čestice,čime je dat uvid u mehanizme kojidominiraju atmosferskom distribucijompolicikličnih aromatičnih ugljovodonika.Metodom dijagnostičkih odnosarealizovana je kvalitativna karakterizacijaizvora emisije PAH u atmosferu.Korekcijom postojećih empirijskih iteorijskih izraza koji opisuju gas/čestičnuparticiju PAH, razvijen je matematičkimodel raspodele policikličnih aromatičnihugljovodonika u ambijentalnom vazduhu.
Gas-particle distribution of 16 EPA primarypolycyclic aromatic hydrocarbons in theatmosphere at the selected sampling sitesin urban, industrial and rural area of theWestern Balkan countries has beenanalyzed within the doctoral dissertation.On the basis of obtained experimentaldata, representative models of transportthrough surface atmospheric layer and gasparticlepartitioning of the semivolatileorganic compounds have been evaluated,giving the overview on mechanisms whichgovern atmospheric distribution ofpolycyclic aromatic hydrocarbons. Usingthe method of diagnostic ratio, qualitativecharacterization of emission sources ofPAHs in the atmosphere has been realized.Through correction of actual empirical andtheoretical expressions, explaining gasparticlepartitioning of PAHs,mathematical model of distribution ofpolycyclic aromatic hydrocarbons in theambient air has been developed.

This thesis is an exloration and exposition of a highly efficient shallow neural network algorithm called word2vec, which was developed by T. Mikolov et al. in order to create vector representations of a language vocabulary such that information about the meaning and usage of the vocabulary words is encoded in the vector space structure. Chapter 1 introduces natural language processing, vector representations of language vocabularies, and the word2vec algorithm. Chapter 2 reviews the basic mathematical theory of deterministic convex optimization. Chapter 3 provides background on some concepts from computer science that are used in the word2vec algorithm: Huffman trees, neural networks, and binary cross-entropy. Chapter 4 provides a detailed discussion of the word2vec algorithm itself and includes a discussion of continuous bag of words, skip-gram, hierarchical softmax, and negative sampling. Finally, Chapter 5 explores some applications of vector representations: word categorization, analogy completion, and language translation assistance.

A falta de alternativas ao modelo normal uni/multivariado já é um problema superado pois atualmente é possível encontrar inúmeros trabalhos que introduzem e desenvolvem generalizações da distribuição normal com relação `a assimetria, curtose e/ou multimodalidade (Branco e Arellano-Valle (2004), Genton (2004), Arellano-Valle et al. (2006)). No contexto dos modelos unidimensionais da Teoria da Resposta ao Item (TRI), Bazán (2005) percebeu esta realidade e introduziu uma classe denominada PANA (Probito Assimétrico - Normal Assimétrica) a qual permite modelar possíveis comportamentos assimétricos de um modelo (uma probabilidade) de resposta ao item bem como a especificação de uma distribuição normal assimétrica para os traços latentes (unidimensionais) a qual é utilizada no processo de estimação. Motivado pela necessidade de melhor representar os fenômenos da área psicométrica (Heinen, 1996, p. 105) e da atual disponibilidade de distribuições elípticas assimétricas cujas propriedades são tão convenientes quanto aquelas devidas `a distribuição normal, a proposta do presente trabalho é apresentar uma extensão do modelo K-dimensional de 3 Parâmetros Probito (Kd3PP) com vetores de traços latentes normalmente distribuídos para o caso t-Assimétrico, gerando, assim, o que denominamos modelo Kd3PP-tA. Nossa proposta, portanto, pode ser considerada como uma extensão do trabalho desenvolvido por Bazán (2005) tanto no sentido de extender a distribuição unidimensional assimétrica dos traços latentes para o caso multidimensional quanto no que conscerne em considerar o achatamento (curtose) da distribuição. Nossa proposta também pode ser vista como uma extensão do trabalho de Béguin e Glas (2001) no sentido de desenvolver o método de estimação bayesiana dos modelos multidimensionais da TRI via DAGS (Dados Aumentados com Amostrador de Gibbs) para o caso em que os vetores de traços latentes comportam-se segundo uma distribuição multivariada t-Assimétrica. No desenvolvimento deste trabalho nos deparamos com uma das principais dificuldades encontradas no processo de estimação e inferência dos modelos multidimensionais da TRI que é a falta de identificabilidade e, com a intenção de ampliar e desmistificar nossos conhecimentos sobre um assunto ainda pouco explorado na literatura da TRI, apresentamos um estudo bibliográfico sobre este tema tanto sob o contexto da inferência clássica quanto bayesiana. Com o intuito de identificar situações particulares em que o uso de uma distribuição normal assimétrica para os traços latentes seja de maior relevância para a estimação e inferência dos parâmetros de item, bem como outros parâmetros relacionados à distribuição dos traços latentes, algumas análises sobre conjuntos de dados simulados são desenvolvidas. Como conclusão destas análises, podemos dizer que há uma melhora superficial quando a informação sobre uma possível assimetria na distribuição dos traços latentes não é ignorada. Além disso, os resultados favoreceram a seleção dos modelos que consideram distribuições assimétricas para os traços latentes, principalmente quando são considerados os modelos que possibilitam a estimação dos parâmetros de localização e escala da distribuição dos vetores de traços latentes. Duas principais contribuições que consideramos de ordem prática, são: a análise e a interpretação de testes através da estimação de modelos uni e multidimensionais da TRI que consideram tanto distribuições simétricas quanto assimétricas para os vetores de traços latentes e a disponibilização de uma função escrita em códigos R e C++ para a estimação dos modelos apresentados e desenvolvidos no presente trabalho.
The lack of alternatives to the univariate or multivariate normal model has been already solved because actually it has been possible to find several works that introduce and develop generalizations of the normal distribution in relation to the asymmetry, kurtosis and/or multimodality (Branco e Arellano-Valle (2004), Genton (2004), Arellano-Valle et al. (2006). In the context of unidimensional models of the Item Response Theory (IRT), Baz´an (2005) observed this fact and introduced a class called PANA (Probito Assimétrico - Normal Assimétrica) which allows to take account for asymmetry in the shape of an item response model (probability) and the specification of a skew normal distribution for unidimensional latent traits which is used in the estimation process. Motivated by the need to better represent the phenomenon of psychometric area (Heinen, 1996, p. 105) and the current availability of skew elliptical distributions whose properties are as convenient as those due to normal distribution, the proposal of this work is to provide an extension of multidimensional 3 Parameters Probit model (Kd3PP) where latent traits vectors are normally distributed for the case of Skew-t distribution (Sahu et al., 2003), generating therefore what we call Kd3PP-St model. Our proposal, therefore, can be regarded as an extension of the work of Bazán (2005) in two ways: the first is extending the unidimensional skew normal distribution of latent traits to the multidimensional case and second in the sense to consider the flattening (kurtosis) of this distribution. Our proposal can also be seen as an extension of the work of B´eguin e Glas (2001) in the sense that we develop the Bayesian estimation method of the 3 parameters multidimensional item response model by DAGS (Augmentated Data with Gibbs sampling) for the case where the latent trait vectors behave according to a Skew-t multivariate distribution. In the development of this work we come across one of the main difficulties encountered in the process of estimation and inference of multidimensional IRT models which is the lack of identifiabilitie and, with the intent to demystify and expand our knowledge on a subject still little explored in the literature of the IRT, we present a bibliographical study on this subject both in the context of classical and Bayesian inference. In order to identify particular situations where the use of a skew normal distribution is more relevant to the estimation and inference of item parameters as well as other parameters related to the distribution of latent traits, some analyses on simulated data sets are developed. As results of these analyses, we can say that there is a modest improvement when information about a possible asymmetry in the distribution of latent traits is not ignored. Moreover, the results favored the selection of models that consider asymmetric distributions for latent traits, especially when models that enable the estimation of parameters of location and scale from this distribution are considered. Two main contributions that we consider of pratical interest are: analysis and interpretations of tests using unidimensional and multidimensional IRT models that consider both simetric and skewed distributions for the vectors of latent traits and a function written in R and C++ language program that is made disponible for the estimation of models treated in this work.

abstract: Lexical diversity (LD) has been used in a wide range of applications, producing a rich history in the field of speech-language pathology. However, for clinicians and researchers identifying a robust measure to quantify LD has been challenging. Recently, sophisticated techniques have been developed that assert to measure LD. Each one is based on its own theoretical assumptions and employs different computational machineries. Therefore, it is not clear to what extent these techniques produce valid scores and how they relate to each other. Further, in the field of speech-language pathology, researchers and clinicians often use different methods to elicit various types of discourse and it is an empirical question whether the inferences drawn from analyzing one type of discourse relate and generalize to other types. The current study examined a corpus of four types of discourse (procedures, eventcasts, storytelling, recounts) from 442 adults. Using four techniques (D; Maas; Measure of textual lexical diversity, MTLD; Moving average type token ratio, MATTR), LD scores were estimated for each type. Subsequently, data were modeled using structural equation modeling to uncover their latent structure. Results indicated that two estimation techniques (MATTR and MTLD) generated scores that were stronger indicators of the LD of the language samples. For the other two techniques, results were consistent with the presence of method factors that represented construct-irrelevant sources. A hierarchical factor analytic model indicated that a common factor underlay all combinations of types of discourse and estimation techniques and was interpreted as a general construct of LD. Two discourse types (storytelling and eventcasts) were significantly stronger indicators of the underlying trait. These findings supplement our understanding regarding the validity of scores generated by different estimation techniques. Further, they enhance our knowledge about how productive vocabulary manifests itself across different types of discourse that impose different cognitive and linguistic demands. They also offer clinicians and researchers a point of reference in terms of techniques that measure the LD of a language sample and little of anything else and also types of discourse that might be the most informative for measuring the LD of individuals.
Dissertation/Thesis
Ph.D. Speech and Hearing Science 2011

碩士
國立高雄科技大學
工業工程與管理系
107
Acceptance sampling plan is one of the widely used tools for quality management. It refers to random sampling of pre-specified samples from the inspection batch, testing or verification, and comparing the results of testing or verification with the original inspection standards. Therefore, it is judged whether the delivery inspection batch meets the original inspection standard, and then determines whether the delivery batch is accepted or rejected. Process capability analysis refers to the study of process capability, the analysis of variability in the manufacturing process relative to product specifications and the stability of process output variation. The process capability index (PCI) is a measure of process capability performance. The tool is designed to improve the part of the process, and the customer can also know from the producers process capability whether the required product meets the requirements. This study uses the commonly used quality management tools, process capability indicators and acceptance sampling plans, and uses the process capability Cpm indicators as the required quality level, and meets the characteristics of the Cpm estimated values, and designs a single acceptance sampling of the measured values, and assumes the test with the hypothesis. The method is to find the n value of the sample required to meet the OC curve (AQL, ) and (RQL, ) and determine the critical value C value of the approved batch, and make a table of acceptance sampling parameters to provide the buyer and the seller to perform the acceptance sampling plan. The reference basis for the judgment of the shipment. Keywords：Process Capability Index, Acceptance Sampling Plan, R Language

Objective: To determine if there was an optimal language sampling context for children with hearing impairment; specifically, if any well-documented method of obtaining a language sample was superior to the others in describing the areas of language that are known to serve as a foundation for later literacy development. Participants: Nine children with hearing impairment who used oral language as their primary mode of communication from the University of Tennessee Child Hearing Services clinic were selected to participate in the study. All were from Caucasian families who spoke English as their primary language and with the exception of hearing impairment, none had other documented disorders. Method: Three language samples were taken in an interview, picture description and story retell format during one 50 minute session. Data Analysis: The language samples were analyzed for syntax and morphology, semantic, pragmatic and narrative measures which are preliteracy factors that influence later literacy acquisition. Results: A battery of language samples is needed in order to appropriately access multiple elements of language relating to literacy acquisition of children with hearing impairment. Conclusion: Through the analysis of this study, it has been determined that in order to get a comprehensive view of language in hearing impaired children who use oral language as their primary communication, a battery of language assessments should be used.

Two of the most under-researched and under-served populations in the field of autism spectrum disorder (ASD), minimally to low-verbal (MLV) individuals and children from low-resource countries, would benefit the most from engaging their parents in research and intervention. First, parents’ unique familiarity with their children could be highly advantageous for language assessment providing a more ecologically valid representation of their children’s abilities. Second, parents’ verbal input, known to predict children’s language, is an important avenue to investigate to guide the development of parent-mediated interventions. Natural language samples, which are used in all three dissertation studies, are ideal for assessing expressive language and for analyzing communicative variations in verbal input. In Study 1, I examined the feasibility of parents (N=33) collecting language samples at home from their MLV children/adolescents with ASD (6;6–19;7years) following a semi-structured elicitation protocol, ELSA-A. I predicted that because of parents’ unique familiarity with their children, they will be better at eliciting speech from them. The results supported this prediction. When with their parents, the MLV children/adolescents produced twice as much speech than when with examiners. Parents collected longer ELSA-As but administered fewer of the recommended activities. Therefore, although parents are not as good at following semi-structured assessment protocols, they elicit speech that is more representative of their children’s everyday abilities. In Studies 2 and 3, I compared the parental input to 37 Bulgarian-speaking (2;7–9;10 years) and 37 English-speaking (1;8–4;9 years) children with ASD matched on expressive language. I compared input in terms of quantity and quality, such as lexical diversity and sentence types (Study-2), and in terms of how parents addressed their children, focusing on personal pronouns, names, and kinship terms (Study-3). Based on past research, I hypothesized that input would differ in quality but not quantity. Indeed, parents’ speech differed in sentence types but not in overall amount. Bulgarian parents asked fewer questions but used more statements. As predicted, they also used significantly more ways to address their children because of the structural characteristics of Bulgarian and potentially different discourse practices. These studies lay the foundation for future cross-linguistic and cross-cultural comparisons in ASD.
2023-08-30T00:00:00Z

Thesis (M.Sc.)--York University, 2004. Graduate Programme in Computer Science.
Typescript. Includes bibliographical references (leaves 112-116). Also available on the Internet. MODE OF ACCESS via web browser by entering the following URL: LINK NOT YET AVAILABLE.

Η παρούσα διατριβή έχει ως σκοπό της, πρώτον, την ανάκτηση συντακτικής πληροφορίας (αναγνώριση συμπληρωμάτων ρημάτων, ανάκτηση πλαισίων υποκατηγοριοποίησης (ΠΥ) ρημάτων, αναγνώριση των ορίων και του είδους των προτάσεων) αυτόματα μέσα από ελληνικά και αγγλικά σώματα κειμένων με την χρήση ποικίλων και καινοτόμων τεχνικών μηχανικής μάθησης και, δεύτερον, την θεωρητική περιγραφή της ελληνικής σύνταξης μέσω τυπικών γλωσσολογικών φορμαλισμών, όπως η γραμματική Ενοποίησης και η γραμματική Φραστικής Δομής Οδηγούμενη από τον Κύριο Όρο. Η διατριβή κινήθηκε πάνω στους εξής καινοτόμους άξονες: 1. Η προεπεξεργασία των σωμάτων κειμένων βασίστηκε σε ελάχιστους γλωσσολογικούς πόρους για να είναι δυνατή η μεταφορά των μεθόδων σε γλώσσες φτωχές σε υποδομή. 2. Η αντιμετώπιση του θορύβου που υπεισέρχεται στα δεδομένα εξ αιτίας της χρήσης ελάχιστων πόρων πραγματοποιείται με Μονόπλευρη Δειγματοληψία. Εντοπίζονται αυτόματα παραδείγματα δεδομένων που δεν προσφέρουν στην μάθηση και αφαιρούνται. Τα τελικά δεδομένα είναι πιο καθαρά και η απόδοση της μάθησης βελτιώνεται πολύ. 3. Αποδεικνύεται η χρησιμότητα της εξαχθείσας πληροφορίας. Η χρησιμότητα των συμπληρωμάτων φαίνεται από την αύξηση της απόδοσης της διαδικασίας ανάκτησης ΠΥ με την χρήση τους. Η χρησιμότητα των εξαγόμενων ΠΥ φαίνεται από την αύξηση της απόδοσης ενός ρηχού συντακτικού αναλυτή με την χρήση τους. 4. Οι μέθοδοι εφαρμόζονται και στα Αγγλικά και στα Ελληνικά για να φανεί η μεταφερσιμότητά τους σε διαφορετικές γλώσσες και για να πραγματοποιηθεί μια ενδιαφέρουσα σχετική σύγκριση ανάμεσα στις δύο γλώσσες. Τα αποτελέσματα είναι πολύ ενθαρρυντικά, συγκρίσιμα με, και σε πολλές περιπτώσεις καλύτερα από, προσεγγίσεις που χρησιμοποιούν εξελιγμένα εργαλεία προεπεξεργασίας.
The thesis aims firstly at the acquisition of syntactic information (detection of verb complements, acquisition of verb subcategorization frames (SF), detection of the boundaries and the semantic type of clauses) automatically from Modern Greek and English text corpora with the use of various state-of-the-art and novel machine learning techniques, and, secondly, at the theoretical description of the Greek syntax through formal grammatical theories like Unification Grammar and Head-driven Phrase Structure Grammar. The thesis has been based on the following novel axes: 1. Corpus pre-processing has been limited to the use of minimum linguistic resources to ensure the portability of the presented methodologies to languages that are poorly equipped with resources. 2. Due to the low pre-processing level, a significant amount of noise appears in the data, which is dealt with One-sided Sampling. Examples that do not contribute to the learning process are detected and removed. The final data set is clean and learning performance improves significantly. 3. The importance of the acquired information is proven. The importance of complements is shown by the improvement in the performance of the SF acquisition process after the incorporation of complement information. The importance of the acquired SF lexicon is shown by its incorporation in a shallow syntactic parser and the increase of the performance of the latter. 4. The methods are applied on Modern Greek and on English to show their portability across different languages and to allow for an interesting rough comparison between the two languages. The results are very satisfactory, comparable to, and in some cases better than, approaches utilizing sophisticated resources for pre-processing.

Based on the examiners’ reports of the University of Cambridge on the shortcomings of the students of form V (school certificate level) in Arabic language in the international examinations of year 2008 and 2009 and on the proposals and recommendations therein, the researcher started this research work to address the difficulties faced by students in Arabic language in the grammatical rules when writing essays and in translation from Arabic to English and vice versa in form V (school certificate level) and chose Doha Academy (secondary department) in Mauritius as model. The research work in fact confirmed the errors of students in the grammatical rules after analyzing the examination scripts of the mock examinations run at the school during the year 2009 and 2010 and the researcher came to know about the causes of errors which were also confirmed by means of a questionnaire distributed to the students at this level. The researcher’s interviews with the male and female teachers of Arabic language at School certificate level on the issue were of great benefits and they in fact made mention of the grammatical errors which the researcher came across in the mock examinations’ scripts after close study and analysis. The researcher also considered the importance of the contrastive analysis between Arabic and English due to its effect in learning Arabic as a second language. ط After analysis of the questionnaire and interviews, the researcher came across a number of important issues which he mentioned in details among the results which he concluded. The researcher concluded that most of the errors in grammar are due to the influence of the English language on the Arabic language through linguistic interference, transfer of experience, generalization, as well as errors related to the target language itself like hypercorrection and intralingual phenomenon etc. And that the errors found in translation are caused mainly because of lack of Arabic vocabularies as well as literal translation and a lack of fluency and confidence in using Arabic structures and grammar. Finally, the researcher mentioned a number of recommendations and suggestions which he considers to be a priority and should be executed if we really wish to eliminate difficulties and problems in the process of teaching and learning of the Arabic language at this level and to see the progress of this language in this island.
على (university of Cambridge) بناء على ما قدمّت جامعة كمبرج قصور الطلبة في اللغة العربية في امتحانات كمبرج الدولية للصف الخامس الثانوية للعامين 8002 8002 م، ومن اقتراحات وتوصيات انطلق الباحث في هذا البحث ليتناول الصعوبات التي – يواجهها طلاب اللغة العربية فيما يتعلق بالقواعد عند كتابة المقالات وفي الترجمة من العربية إلى الانجليزية والعكس وذلك في الصف الخامس الثانوي في جمهورية موريشيوس م اختيار أكاديمية الدوحة في موريشيوس نموذجا. لقد أشار البحث فعلا إلى أخطاء الطلاب والطالبات بعد تحليل أوراق إجابات لامتحانين 8000 م كما - في اللغة العربية أُجرِيا في الأكاديمية في العامين 8002 (mock exams) تجريبيين تعرّف الباحث على أسباب الأخطاء التي أكّدها الاستبيان الموزّع على طلاب هذا المستوى. وكانت لمقابلات الباحث م مدرسي ومدرسات هذا المستوى حول القضية ثمراتها حيث أكّدت كذلك ما وجدها الباحث من الأخطاء في النحو والصرف في أوراق إجابات الأمتحان بعد الدراسة والتحليل. رأى الباحث كذلك أهمية التحليل التقابلي بين العربية والانجليزية لتأثير هذه الأخيرة في تعلّم الطلبة اللغة العربية كلغة ثانية. بعد تحليل الاستبيان والمقابلات وصل الباحث إلى عدد من القضايا المهمة وذكرها بالتفصيل في النتائج في آخر البحث. وصل الباحث إلى أن اكثر الأخطاء ترج إلى تأثير اللغة الانجليزية على اللغة العربية تتمثل في التداخل اللغوي ونقل الخبرة والتعميم السلبي كما أن هناك أخطاء تتعلق باللغة اهمدف كالمبالغة في التصويب وتداخل اللغة نفسها. وترج الأخطاء في الترجمة أساسا إلى النقص في المفردات العربية، فضلا عن الترجمة الحرفية وعدم الطلاقة والثقة في استخدام التراكيب العربية والتدقيق النحوي. وأخيرا ذكر الباحث عددا من التوصيات والاقتراحات التي يعتبرها من الأولويات والضروريات ا ولتي يوصي بتنفيذها إذا أُريد بالفعل القضاء على الصعوبات والمشكلات في تعليم اللغة العربية وتعلمه وتقدم هذه اللغة في هذه الجزيرة.
Religious Sudies & Arabic
D.Litt. et Phil. (Arabic)

This research aimed to determine whether the rejection by Russian Orthodox Church leaders of recent translations of the Bible into Russian could be ascribed to a conflict of Russian and Western translation norms. Using Lefevere's (1992) notion of systems, the study compared the norms of Russian Bible translations, Western Bible translation and Russian literary translation, as well as those of a segment of the target audience, to determine the extent of their compatibility with each other and with the translations in question. The results showed that the recent translations did reflect the norms of Western Bible translation, but that these were not atypical of norms for previous Russian and Slavonic translations, nor for the norms of Russian literary translation. However, the results also showed that in practice target audience norms mirrored those of the Russian Orthodox Church, resulting in a similar rejection of the newer translations.
Linguistics
M.A. (Linguistics)

Programmers increasingly rely on managed languages (e.g. Java and C#) to develop applications faster and with fewer bugs. Managed languages encourage allocating objects in the heap and rely on automatic memory management (garbage collection) to reclaim objects the program can no longer access. With more objects in the heap, the heap encodes more program state than ever before and offers new opportunities for optimization and analysis. This dissertation shows how to efficiently leverage the managed runtime to perform dynamic heap analysis. Previous heap analysis approaches significantly slow down programs, require special hardware, and/or increase memory consumption by 75% or more. We presents two synergistic techniques—dynamic object sampling (DOS) and heap summarization (HSG)—that mine program state embedded in the heap efficiently enough to use in production and effectively enough to improve performance, find bugs, and increase program understanding. We use these techniques to address three problems: (1) Performance of managed language. Because some objects live for a long time, they incur disproportionate collection costs. We optimize these costs with dynamic pretenuring. Dynamic pretenuring uses DOS to accurately predict allocation site survival rates and uses these predictions to improve performance. (2) Finding bugs. Memory leaks in managed languages occur when a program inadvertently maintains references to objects that it no longer needs. Along with degrading performance and resulting in program crashes, memory leaks cause systematic heap growth. We introduce Cork which uses the simplest type of HSG, a class points-from summary graph (CPFG), to detect systematic heap growth. Cork quickly identifies growing data structures observed in three popular benchmarks (fop, jess, and jbb2000) while adding an average of only 2.3% to total time. Additionally, we use Cork to debug a reported memory leak in Eclipse. (3) Program understanding. For a long time, static analysis has sought to statically summarize the shape of dynamic data structures to aid in program verification and understanding. Unfortunately, it only works on small programs. We introduce ShapeUp which instead characterizes recursive data structures dynamically by discovering data structure shape and degree invariants at runtime. ShapeUp uses DOS and a class field-wise summary graph (CFSG) to track in- and out-degree invariants of data structure nodes. We show how ShapeUp automatically identifies recursive data structures and likely shape invariants. Finally, we monitor discovered shape invariants to detect when a data structure becomes malformed. In summary, this dissertation is the first to leverage the managed runtime to perform dynamic heap analysis both accurately and efficiently. Our results show that the heap contains an enormous amount of program state and that there is much potential for dynamically mining heap characteristics for optimization, debugging, and program understanding.
text

The general view of educators at all levels of education as well as trainers in industry, that learners do not have the ability to use their knowledge inter-curricular and holistically in their lives prompted the study. A very superficial study that was done in 2000, regarding the same problem, acted as the starting point for this study. The main purpose of the study was to determine the reasons for the general perception by educators that learners do not use their embedded knowledge in an inter-curricular manner. It also aimed to determine the most effective facilitating styles and methods that would help the learners to develop the skills to learn with the main aim to achieve learning that lasts and not just learning to pass. The use of deconstruction as a learning activity forms the biggest part of this qualitative action research project which was based on the main principles of the hermeneutic approach namely, communication followed by interpretation in order to reach understanding. The principles of narrative therapy were also used in the project. The hermeneutic approach followed in this study also allows the reader to walk the road with the participants as a narrative, first person reporting style provides rich background information regarding all the participants of the study as well as the contexts in which the actions took place. The reader will also be able to feel the closeness that existed between the educator and the learners who were part of this study.
Educational Studies
M.Ed. (Didactics)