No description available.

Sentence labeling is a type of pattern recognition task that involves the assignment of a categorical label to each member of a sentence of observed words. Standard supervised sentence-labeling systems often have poor generalization: it is difficult to estimate parameters for words which appear in the test set, but seldom (or never) appear in the training set, because they only use words as features in their prediction tasks. Representation learning is a promising technique for discovering features that allow a supervised classifier to generalize from a source domain dataset to arbitrary new domains. We demonstrate that features which are learned from distributional representations of unlabeled data can be used to improve performance on out-of-vocabulary words and help the model to generalize. We also argue that it is important for a representation learner to be able to incorporate expert knowledge during its search for helpful features. We investigate techniques for building open-domain sentence labeling systems that approach the ideal of a system whose accuracy is high and consistent across domains. In particular, we investigate unsupervised techniques for language model representation learning that provide new features which are stable across domains, in that they are predictive in both the training and out-of-domain test data. In experiments, our best system with the proposed techniques reduce error by as much as 11.4% relative to the previous system using traditional representations on the Part-of-Speech tagging task. Moreover, we leverage the Posterior Regularization framework, and develop an architecture for incorporating biases from prior knowledge into representation learning. We investigate three types of biases: entropy bias, distance bias and predictive bias. Experiments on two domain adaptation tasks show that our biased learners identify significantly better sets of features than unbiased learners. This results in a relative reduction in error of more than 16% for both tasks with respect to existing state-of-the-art representation learning techniques. We also extend the idea of using additional unlabeled data to improve the system's performance on a different NLP task, word alignment. Traditional word alignment only takes a sentence-level aligned parallel corpus as input and generates the word-level alignments. However, as the integration of different cultures, more and more people are competent in multiple languages, and they often use elements of multiple languages in conversations. Linguist Code Switching (LCS) is such a situation where two or more languages show up in the context of a single conversation. Traditional machine translation (MT) systems treat LCS data as noise, or just as regular sentences. However, if LCS data is processed intelligently, it can provide a useful signal for training word alignment and MT models. In this work, we first extract constraints from this code switching data and then incorporate them into a word alignment model training procedure. We also show that by using the code switching data, we can jointly train a word alignment model and a language model using co-training. Our techniques for incorporating LCS data improve by 2.64 in BLEU score over a baseline MT system trained using only standard sentence-aligned corpora. / Computer and Information Science

Having a linear representation of magnitude across scales is essential in understanding many scientific concepts (Tretter, et al., 2006a) and is predictive of a range of mathematical achievement tests (Siegler & Booth, 2004). Despite the importance of understanding magnitude and scale, people have substantial difficulty comparing magnitudes outside of human perception (e.g., Jones, et al., 2008). The present work aims to examine the way people learn to represent and reason about large magnitudes through the development of two science of learning activities based on hierarchical alignment activity and corrective feedback. The hierarchical alignment activity utilizes several analogical reasoning principles: hierarchical alignment, progressive alignment, structural alignment, and multiple opportunities to make analogies. Study 1 examines the effectiveness of hierarchical alignment by contrasting it with a conventional activity that uses all the analogical reasoning principles described above except for hierarchical alignment. Study 2 examines a corrective feedback activity, based on the same analogical reasoning principles used in study 1, except, using corrective feedback instead of progressive alignment and hierarchical alignment. Thus, study 2 examines the necessity of hierarchical and progressive alignment. That both activities were successful in developing linear representations of geologic time (and for study 1, astronomical distances), suggests that multiple opportunities to make analogies through structural alignment are key components in developing analogies for learning magnitude. There appears to be an additive benefit of including hierarchical alignment (i.e., practice aligning magnitude relations across scales) in analogies for learning about magnitudes. Corrective feedback may also be a useful strategy in learning about scale information. Pedagogical implications are discussed. Both activities were based on the hypothesis that magnitudes at scales outside human perception are represented and reasoned about in the same way as magnitudes at human scales. The Category Adjustment Model (Huttenlocher, et al., 1988) suggests magnitude at human scales is stored as a hierarchical combination of metric and categorical information. People may use category boundaries to help make estimations in lieu of precise metric information. Variation in estimation, therefore, occurs because of imprecision of category boundaries (Shipley & Zacks, 2008; Zacks & Tversky, 2001). The current studies provided salient category boundaries to develop a more linear representation of magnitude. Thus, the effectiveness of the hierarchical alignment activity and the corrective feedback activity supports the hypothesis that people use hierarchically organized categorical information when making estimations across scales and across dimensions; and that providing people with more salient category boundary information improves estimation. Similarities and differences among temporal, spatial, and abstract line estimations are identified. Theoretical implications, including the potential application of the Category Adjustment Model to mental number lines, are discussed. / Psychology

This study is an investigation of (dis)alignment and (dis)affiliation in narrative conversations in English. The analytical focus is on the teller's approaches for eliciting an assessment from the recipient, and the recipient's questioning behaviors to request further information. Conversation analysis (CA) is employed to analyze and interpret data from fourteen naturally occurring, mundane conversations between American and Japanese participants. Three conversations were audio-taped only, and eleven were both audio- and video-taped. All recorded conversations were transcribed using CA conventions. The analysis of the data revealed the teller's approaches for eliciting an assessment from the recipient, with repetitions being most frequently observed. Gazing, nodding, and using gestures, and multiple approaches including elaborating on the story, repeating a point, and emphasizing important words in a louder voice, were also observed. The analysis also revealed that minimal responses from the recipient such as "I see" did not necessarily indicate the recipient's understanding of the story. The recipient's assessments and comments functioned as a barometer of the degree of understanding. Several types of recipient questioning behaviors were identified, including initiating repair, requesting background information, prompting further story, expressing disagreement, and assisting the teller to continue the story. These questions indicated the degree to which the recipient had understood the story, encouraged the teller to proceed with the story, satisfied the recipient's personal curiosity, showed the recipient's interest and involvement in the story, challenged the teller's interpretation of the story, and assisted the teller when in trouble. Whereas questions sometimes caused temporal derailment, they could also contribute to enhancing alignment and affiliation between the participants. Whether or not a particular question was an indication of (dis)alignment or (dis)affiliation depended on which aspect of the story the teller and the recipient focused on. Although the smooth flow of narratives might be preferable, in naturally occurring mundane conversations, the recipient sometimes needs to intervene in the teller's talk. What seems important in communication is to attain shared understanding in the end, even if the intervention delays the development of the narrative. / CITE/Language Arts

Background: Lower limb amputees face many challenges, and most of them prefer to use prosthetics for daily tasks and activities. The prosthesis is usually a combination of connected prosthetic components, and their spatial orientation is called the prosthetic alignment. Proper alignment is essential, as it substantially affects the quality and comfort of a prosthesis.   Objective: The aim of this study was to create a method that could accurately and effectively quantify the alignment of a transtibial and transfemoral prostheses using Vicon optical motion capture system.   Methods: Two experimental series were conducted. The first one was to test the repeatability of the measurement. Three analysts placed retroreflective markers on the prostheses three times, and five measurements were recorded each time. Alignment parameters were calculated in Vicon ProCalc for each measurement, and a standard error of measurement was found for each alignment parameter. The standard error of measurement was calculated from three variance components, between-analyst, within-analyst, and between-trial variability. The second experimental series was conducted to understand the relationship between alignment adjustments and the outcome parameters. The socket height, internal rotation, flexion, adduction, and translation were modified and measured. The socket translation was calculated in three coordinate systems to study how they affect the outcome.   Results: For the first experimental series, the standard error of measurement for every alignment parameter was below 3° and 6 mm. The between-analyst variability was the most prominent, and the parameters calculated in the sagittal plane were more reliable than those calculated in the frontal and transverse plane. In the second experimental series, there was a linear relationship between the modifications and the measured outcome. When a connection between two prosthetic components was changed by turning the screws one round, the average change in angle between them was 2°, and the average translation change was 4.4 mm. Of the three coordinate systems, the translation calculated in ankle coordinates was more reliable than in global coordinates and describe the translation more effectively than in socket coordinates.   Conclusion: The reliability of the measurements was considered good. The standard error of measurement was low, and the main variability resulted from differences in marker placement between the analysts. The results from the measured alignment changes were as expected. All the parameters could be effectively interpreted, and the ankle coordinates were considered advantageous in describing the socket translation.

The advent of multi-photon microscopes and highly sensitive protein sensors enables the recording of astrocyte activities on a large population of cells over a long-time period in vivo. Existing tools cannot fully characterize these activities, both within single cells and at the population-level, because of the insufficiency of current region-of-interest-based approaches to describe the activity that is often spatially unfixed, size-varying, and propagative. Here, we present Astrocyte Quantitative Analysis (AQuA), an analytical framework that releases astrocyte biologists from the ROI-based paradigm. The framework takes an event-based perspective to model and accurately quantify the complex activity in astrocyte imaging datasets, with an event defined jointly by its spatial occupancy and temporal dynamics. To model the signal propagation in astrocyte, we developed graphical time warping (GTW) to align curves with graph-structured constraints and integrated it into AQuA. To make AQuA easy to use, we designed a comprehensive software package. The software implements the detection pipeline in an intuitive step by step GUI with visual feedback. The software also supports proof-reading and the incorporation of morphology information. With synthetic data, we showed AQuA performed much better in accuracy compared with existing methods developed for astrocytic data and neuronal data. We applied AQuA to a range of ex vivo and in vivo imaging datasets. Since AQuA is data-driven and based on machine learning principles, it can be applied across model organisms, fluorescent indicators, experimental modes, and imaging resolutions and speeds, enabling researchers to elucidate fundamental astrocyte physiology. / Doctor of Philosophy / Astrocyte is an important type of glial cell in the brain. Unlike neurons, astrocyte cannot be electrically excited. However, the concentrations of many different molecules inside and near astrocytes change over space and time and show complex patterns. Recording, analyzing, and deciphering these activity patterns enables the understanding of various roles astrocyte may play in the nervous system. Many of these important roles, such as sensory-motor integration and brain state modulation, were traditionally considered the territory of neurons, but recently found to be related to astrocytes. These activities can be monitored in the intracellular and extracellular spaces in either brain slices and living animals, thanks to the advancement of microscopes and genetically encoded fluorescent sensors. However, sophisticated analytical tools lag far behind the impressive capability of generating the data. The major reason is that existing tools are all based on the region-of-interest-based (ROI) approach. This approach assumes the field of view can be segmented to many regions, and all pixels in the region should be active together. In neuronal activity analysis, all pixels in an ROI (region of interest) correspond to a neuron and are assumed to share a common activity pattern (curve). This is not true for astrocyte activity data because astrocyte activities are spatially unfixed, size-varying, and propagative. In this dissertation, we developed a framework called AQuA to detect the activities directly. We designed an accurate and flexible detection pipeline that works with different types of astrocyte activity data sets. We designed a machine learning model to characterize the signal propagation for the pipeline. We also implemented a compressive and user-friendly software package. The advantage of AQuA is confirmed in both simulation studies and three different types of real data sets.

Real-time face recognition has recently become available for the government and industry due to developments in face recognition algorithms, human head detection algorithms, and faster/low cost computers. Despite these advances, however, there are still some critical issues that affect the performance of real-time face recognition software. This paper addresses the problem of off-centered and out-of-pose faces in pictures, particularly in regard to the eigenface method for face recognition. We first demonstrate how the representation of faces by the eigenface method, and ultimately the performance of the software depend on the location of the eyes in the pictures. The eigenface method for face recognition is described: specifically, the creation of a face basis using the singular value decomposition, the reduction of dimension, and the unique representation of faces in the basis. Two different approaches for aligning the eyes in images are presented. The first considers the rotation of images using the orthogonal Procrustes Problem. The second approach looks at locating features in images using energy-minimizing active contours. We then conclude with a simple and fast algorithm for locating faces in images. Future research is also discussed. / Master of Science

[EN] EHealth plays an essential role in supporting healthcare in today's digital society; it is perceived as crucial for high quality and cost-effective healthcare. However, getting the expected benefits from eHealth has been difficult to demonstrate. There has been a raising interest in adopting eHealth Governance frameworks to obtain re-assurance that investments return the expected results in health care. How IT Governance is implemented within healthcare, the actual impact on strategic alignment and its influence to the information society progress, remains poorly understood. For this purpose we have explored the application of these frameworks within the National Health Service in Scotland and their impact on the following three aspects: eHealth Governance maturity, strategic alignment with healthcare and local progress of digital societies. This research is a longitudinal study (2008-2013), involving an exploratory and explanatory multi-case analysis of three representative organisations across Scotland. A combination of empiric methods has been used: semi-structured interviews with implementers, surveys (Strategic Alignment Model), cross-sectoral/national benchmarking based on a literature review and a qualitative analysis of established eHealth progress indicators. Ninety-two participants have been involved across three case studies. The outcomes of this study have been published over a period of 5 years representing a composite thesis based on relevant publications. Results sustain that EHealth Governance is in its infancy across sectors and countries. 80% of the organisations worldwide are in a transition point between a "committed" and an "established" process. Our results support that the more mature eHealth Governance is, the better the strategic alignment between eHealth and health care organisations (HCOs), hence the better progress of eHealth and the Digital Society. The Strategic alignment is slowly maturing across organisations (15% since 2008), indicating a faster development than the overall Digital Society (Scotland) progress indicators. The National eHealth Strategy shows signs of steady progress and very positive eHealth uptake in society with an overall growth of 12% since 2008, despite the deep economical recession within the period of this research. The conclusions of this study as a longitudinal analysis are limited and more research over the forthcoming years is required. For this purpose, a simplified and adapted method to monitor these trends in future HCOs research has also been provided. / [ES] La eSalud juega un papel esencial en el desarrollo de la asistencia médica en sociedades digitales; se percibe como un elemento crucial en la provisión de servicios médico-sanitarios alta calidad y costo-efectivos. A pesar de ello, hasta ahora ha sido difícil demostrar la materialización de los beneficios esperados de la eSalud, pero hay un interés creciente en la adopción de marcos de referencia basados en buenas prácticas, y estándares profesionales internacionales para la gestión y dirección de la eSalud, con el propósito de asegurar que las inversiones revierten los resultados esperados en el cuidado y servicios de la salud. Hasta ahora se sabe muy poco sobre el fenómeno de cómo la eSalud es integrada en el cuidado y servicios de la salud, y del impacto que esta tiene en la alineación estratégica de la eSalud. Igualmente, sabemos muy poco de la influencia real que estas prácticas tienen en el progreso de sociedades digitales. Este estudio se centra en explorar la aplicación de buenas practicas y estándares internacionales como marco de referencia en el gobierno de la eSalud en el servicio de salud Escocés; también contrastamos y comparamos el fenómeno con otros países y sectores. Esta investigación es un estudio longitudinal (2008-2013) que incorpora un análisis exploratorio y explicativo de casos. Se obtuvieron un total de noventa y dos participantes a lo largo de los tres casos estudiados, con representación de los principales grupos de interés (médicos y no médicos). Los resultados se han divulgado a lo largo del periodo de investigación en un compendio de publicaciones relevantes que conforman la tesis. Los principales hallazgos muestran que el gobierno de la eSalud está en su infancia en los sectores y países analizados: el 80% de las organizaciones a nivel mundial presentan este proceso en un punto de transición entre "comprometido" y "establecido" (Modelo SAM). Los resultados corroboran que cuanto más maduro es el gobierno de la eSalud, mayor alineación estratégica entre la eSalud y la organización, y mayor progreso de la variable eSalud en los indicadores de la sociedad de la información. La alineación estratégica esta madurando lentamente (15% desde 2008), sin embargo este crecimiento es mas rápido que el progreso observado en los indicadores de la sociedad digital (Escocia). La estrategia nacional Escocesa muestra signos de progreso sostenido y de integración (por uso o adopción) de las iniciativas de eSalud en la sociedad (crecimiento del 12% desde 2008), a pesar de la profunda depresión económica durante el periodo de investigación. Las conclusiones de esta investigación, como estudio longitudinal, son limitadas y requieren la captura de más datos y observaciones durante los próximos anos. Con el fin de facilitar este proceso, se ha propuesto un método simplificado y adaptado al sector salud, que permite capturar observaciones, comparar y monitorizar estas tendencias en futuras investigaciones en el sector salud. / [CA] La eSalut juga un paper essencial en el suport a l'assistència sanitària a la societat digital de hui en dia; es percep com crucial per a l'alta qualitat i efectivitat del servicis de salut. No obstant això, ha estat difícil de demostrar la obtenció dels beneficis esperats de la eSalut. Hi ha hagut un interès en augmentar l'adopció de marcs de governança de la eSalut per obtenir re-assegurament que les inversions retornen els resultats esperats en els servicis sanitaris. Com s'implementa la governança de les TIC dins de l'assistència sanitària, l'impacte real en l'alineació estratègica i la seua influència en el progrés de la societat de la informació, continua sent poc conegut. Aquest estudi explora l'aplicació d'aquests estàndards i marcs de referència dins dels Serveis Nacionals de Salut d'Escòcia i el seu impacte en els tres aspectes següents: la maduresa de la governança de la eSalut", l'alineació estratègica amb l'assistència sanitària i, finalment, el progrés local de les societats digitals. Aquesta investigació és un estudi longitudinal (2008-2013), que implica una anàlisi multi cas exploratori i explicatiu de tres organitzacions representatives del servici nacional de salut de Escòcia. S'ha utilitzat una combinació de mètodes empírics: entrevistes semi estructurades, enquestes (Model SAM) comparatives de mercat intersectorial i internacional basat en una revisió bibliogràfica i, finalment, una anàlisi qualitativa dels indicadors de progrés eSalut. Noranta dos participants han informat a través de tres estudis de casos. Els resultats s'han divulgat al llarg del període d'investigació en un compendi de publicacions rellevants que conformen la tesi. Els resultats assenyalen que la governança de la eSalut està en la seua infància en tots els sectors i països. 80% de les organitzacions de tot el món es troben en un punt de transició entre un procés "compromès" i "establint". Els nostres resultats apunten que quan més madur es la governança de la eSalut, millor serà l'alineació estratègica entre la eSalut i les organitzacions d'atenció sanitària (HCOs), per tant el millor progrés de la sanitat electrònica a la Societat Digital. L'alineació estratègica està madurant lentament en les organitzacions (15% des de 2008); aquest desenvolupament és més ràpid que el progrés de la societat digital (Escòcia). L'estratègia Nacional de eSalut mostra signes de progrés constant i l'absorció de la eSalut en la societat es prou positiva, amb un creixement global del 12% des de l'any 2008, tot i la profunda recessió econòmica durant el període de temps d'aquesta investigació. Les conclusions d'aquest estudi com una anàlisi longitudinal són limitades i es requereix més investigació en els propers anys. / Beratarbide Sobrado, ME. (2016). Longitudinal analysis of eHealth Governance within healthcare organizations as a critical factor in the adaptation to the Information Society in Scotland [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/63456

Precisely positioned optical lenses are currently required for many highly repetitive mechanics and applications. Thus the need for micron-scale repetition between opto-mechanical units is evident, especially in industrial manufacturing and medical breakthroughs. In this thesis, a novel optical metrology system is proposed, designed, and built whose purpose is to precisely locate the center of a mechanical fixture and then to assemble a plano-convex optical lens into the located position of the fixture. Center location specifications up to ±3 µm decenter and ±0.001° tilting accuracy are required. Nine precisely positioned lenses and fixtures were built with eight units passing the requirements with a repetitive standard deviation of ±0.15 µm or less. The assembled units show satisfactory results.

Yes / Companies in the high technology manufacturing and development sector have to continually innovate in order to survive and grow in increasingly turbulent and competitive markets. It is common practice for the parent company to spin off separate business units that can incubate and capitalise on the development of new technological innovations in order to grow and create new markets. This case study illustrates the issues that arise when a separate “maverick” business unit focusing on developing a new and disruptive innovation is spun off from the parent company. It underlines the problems that arise when ICT systems and operational processes are not strategically aligned and imposed by the parent company. It also demonstrates how innovative business units can harness their unique talents and apply them to solving operational problems. By developing a new bespoke system aligned with the maverick unit’s emergent processes, the maverick business unit was pulled back from the brink of disaster to a successful and profitable business unit.