No description available.

Customers adoption of Internet banking has become a widely-researched topic, although it is fair to state that some research gaps still exist. This research aims to fill some of the research gaps by examining the factors that determine the relevant behaviour of three different categories of Internet banking customers in China (i.e. current users, non-users, and discontinued users), and by developing two conceptual models that are derived from different, but complementary, theoretical approaches. The Decision Making Model and the Service and Relationship Evaluation Model are developed in this research. The Decision Making Model is grounded in the technology acceptance model (TAM) and it incorporates an additional construct of perceived value of using Internet banking. Additionally, the Service and Relationship Evaluation Model is derived from the service quality evaluation and relationship quality evaluation literature. Unlike in most other Internet banking adoption studies, these two conceptual models are used complementarily to deliver a comprehensive understanding of customers Internet banking adoption in China. The models are tested using a sample of 614 Chinese Internet banking customers collected via mall-intercept personal interviews based on questionnaires. Partial Least Square (PLS) path modelling and mediation analysis are applied to test the hypotheses advanced in the two models. The key findings of this research show that perceived value is a major factor for explaining customers Internet banking adoption, thus indicating to the banks that they should reduce costs associated with using Internet banking while providing more (perceived) benefits to customers; the importance of incorporating perceived value in Internet banking adoption model(s) is also demonstrated. The findings also confirm that perceived usefulness and perceived ease of use are important factors that determine the adoption of Internet banking by all categories of customers. Current users and non-users perceptions of their behavioural control over using Internet banking contribute to their adoption of Internet banking, and such control perceptions are shaped by self-efficacy, perceived government support and technological support. Additionally, it is demonstrated that both current users and discontinued users perceived value and perceived service quality of Internet banking have positive associations with their satisfaction with Internet banking, which lead to their Internet banking adoption. Moreover, the findings reveal that current users are more likely to continue with Internet banking if they are affectively committed to their banks; they are less likely to continue with Internet banking if they are calculatively committed to their banks due to the costs associated with leaving the banks. These therefore indicate the importance of establishing high-quality customer-bank relationships and placing less strict switching cost barriers that impose less pressure on their existing customers. This research contributes to the Internet banking adoption literature by (i) identifying the important category of Internet banking discontinued users, apart from current users and non-users; and (ii) using two complementary conceptual models, which are grounded in different theoretical streams, to investigate the relevant adoption behaviour of all three categories of Internet banking customers. It hence delivers a comprehensive understanding of personal customers adoption of Internet banking in China.

With the increasing use of the Internet several problems have arisen regarding the processing of electronic documents. These include content filtering, content retrieval/search. Moreover, document security has taken a centre stage including copyright protection, broadcast monitoring etc. There is an acute need of an effective tool which can find the identity, location and the time when the document was created so that it can be determined whether or not the contents of the document were tampered with after creation. Owing the sensitivity of the large amounts of data which is processed on a daily basis, verifying the authenticity and integrity of a document is more important now than it ever was. Unsurprisingly document authenticity verification has become the centre of attention in the world of research. Consequently, this research is concerned with creating a tool which deals with the above problem. This research proposes the use of a Quick Response Code as a message carrier for Text Key-print. The Text Key-print is a novel method which employs the basic element of the language (i.e. Characters of the alphabet) in order to achieve authenticity of electronic documents through the transformation of its physical structure into a logical structured relationship. The resultant dimensional matrix is then converted into a binary stream and encapsulated with a serial number or URL inside a Quick response Code (QR code) to form a digital fingerprint mark. For hiding a QR code, two image steganography techniques were developed based upon the spatial and the transform domains. In the spatial domain, three methods were proposed and implemented based on the least significant bit insertion technique and the use of pseudorandom number generator to scatter the message into a set of arbitrary pixels. These methods utilise the three colour channels in the images based on the RGB model based in order to embed one, two or three bits per the eight bit channel which results in three different hiding capacities. The second technique is an adaptive approach in transforming domain where a threshold value is calculated under a predefined location for embedding in order to identify the embedding strength of the embedding technique. The quality of the generated stego images was evaluated using both objective (PSNR) and Subjective (DSCQS) methods to ensure the reliability of our proposed methods. The experimental results revealed that PSNR is not a strong indicator of the perceived stego image quality, but not a bad interpreter also of the actual quality of stego images. Since the visual difference between the cover and the stego image must be absolutely imperceptible to the human visual system, it was logically convenient to ask human observers with different qualifications and experience in the field of image processing to evaluate the perceived quality of the cover and the stego image. Thus, the subjective responses were analysed using statistical measurements to describe the distribution of the scores given by the assessors. Thus, the proposed scheme presents an alternative approach to protect digital documents rather than the traditional techniques of digital signature and watermarking.

In this thesis we investigate several methods how to improve the quality of statistical machine translation (MT) by using linguistically rich information. First, we describe SemPOS, a metric that uses shallow semantic representation of sentences to evaluate the translation quality. We show that even though this metric has high correlation with human assessment of translation quality it is not directly suitable for system parameter optimization. Second, we extend the log-linear model used in statistical MT by additional source-context model that helps to better distinguish among possible translation options and select the most promising translation for a given context.

Estudos indicam que a utilização de aplicativos educacionais móveis vêm crescendo continuamente, possibilitando a alunos e professores maior flexibilidade e comodidade na execução de atividades e práticas educacionais. Embora várias instituições já tenham aderido à modalidade de aprendizagem móvel (m-learning), sua adoção ainda traz problemas e desafios organizacionais, culturais e tecnológicos. Um destes problemas consiste em como avaliar adequadamente a qualidade dos aplicativos educacionais desenvolvidos. De fato, os métodos existentes para avaliação da qualidade de software ainda são muito genéricos, não contemplando aspectos específicos aos contextos pedagógico e móvel. Nesse cenário, o presente trabalho apresenta o método MoLEva, desenvolvido para avaliar a qualidade de aplicativos educacionais móveis. O método tem como base a norma ISO/IEC 25000, sendo composto por: (i) modelo de qualidade; (ii) métricas; e (iii) critérios de julgamento. Para validar o método, foram realizados dois estudos de caso; o primeiro consistiu na aplicação do MoLEva para avaliar o aplicativo do ENEM; o segundo consistiu na aplicação do método para avaliação de aplicativos para o ensino de idiomas. A partir dos resultados obtidos, foi possível identificar problemas e pontos de melhoria nos aplicativos avaliados. Além disso, os estudos de caso conduzidos forneceram bons indicativos a respeito da viabilidade de uso do método MoLEva na avaliação de aplicativos educacionais móveis. / Studies indicate that the use of mobile learning applications has grown continuously, allowing students and teachers greater flexibility and convenience in the execution of educational activities and practices. Although several institutions have already adhered to the mobile learning (m-learning) modality, their adoption still brings organizational, cultural and technological problems and challenges. One of these problems is how to adequately evaluated the quality of the mobile learning applications developed. In fact, existing methods for evaluating software quality are still very generic, not considering aspects specific to the pedagogical and mobile contexts. In this scenario, the present work presents the MoLEva method, developed to evaluate the quality of mobile learning applications. The method is based on the ISO / IEC 25000 standard, being composed of: (i) quality model; (ii) metrics; and (iii) criteria of judgment. To validate the method, two case studies were performed; the first consisted of applying MoLEva to evaluate the ENEM application; the second consisted of applying the method for evaluating applications for language teaching. From the obtained results, it was possible to identify problems and improvement points in the evaluated applications. In addition, the case studies conducted provided good indications regarding the feasibility of using the MoLEva method in evaluating mobile learning applications.

A qualidade dos serviços desenvolvidos por uma instituição de saúde depende muito da competência técnica e da habilidade de interação e comunicação de seus trabalhadores para com o usuário. As intervenções técnicas realizadas pela equipe de enfermagem requerem avaliações permanentes face aos riscos que comportam. Este estudo, observacional e seccional, analisou a qualidade e o tempo de execução dos procedimentos, banho e curativo, realizados pela equipe de enfermagem em pacientes internados na unidade médico-cirúrgica de um Hospital Universitário no estado do Paraná, tomando por referência o grau de dependência assistencial desta clientela. A população alvo foi constituída dos seguintes procedimentos: banho, em três tipologias (aspersão, aspersão com auxílio de cadeira de banho e banho no leito) e curativos executados em pacientes classificados segundo grau de dependência da assistência de enfermagem (I, II, III e IV). A amostragem para ambos os procedimentos foi por conveniência. Os dados foram coletados por meio de observação direta com a utilização de um instrumento tipo check list (lista de verificação). A qualidade dos procedimentos foi analisada com base no Índice de Positividade (IP) para cada item do instrumento e no escore de acertos, obtidos em cada procedimento; o tempo de execução foi avaliado em minutos. A qualidade da execução do procedimento foi considerada satisfatória quando a mediana do IP e do escore mediano de acertos fosse ? 70%. Foram observados 258 banhos de aspersão (42,6% grau I, 42,6% grau II e 14,8% grau III); 98 de aspersão com auxílio de cadeira de banho (12,5% grau I, 26,5% grau II, 54,1% grau III e 7,1% grau IV) e 46 banhos no leito (4,3% grau I, 37% grau III e 58,7% grau IV). O IP superou 70% apenas no banho de aspersão com auxílio de cadeira de banho em pacientes classificados no grau IV e no banho no leito no grau I. Os itens mais comprometidos estão relacionados à orientação/comunicação/interação com o paciente, higiene oral, desinfecção concorrente do leito, inspeção das condições da pele e valorização das queixas do paciente. Os resultados obtidos nos scores medianos de acertos nas três tipologias de banho, nos quatro graus de dependência também evidenciaram baixa qualidade, uma vez que somente no banho de aspersão com auxílio de cadeira de banho, grau IV e banho no leito grau I, 50% dos procedimentos alcançaram scores de até 80% e 76,5% de acertos, respectivamente. No procedimento curativo dos 168 observados, 33,9% foram em pacientes de grau I, 38,7% de grau II, 19,6% de grau III e 7,8% de grau IV. Em todos os graus de dependência, alcançou-se o índice de positividade total recomendado (? 70%). Entretanto, itens como preparo adequado do ambiente, conferência do prazo de validade dos materiais, respeito aos princípios de assepsia e manutenção da seqüência lógica do procedimento mostram baixa positividade. Os scores medianos de acertos foram superiores a 70% em todos os graus de dependência, indicando que o procedimento atende um padrão de qualidade. Tanto nos banhos como nos curativos não foram observadas diferenças no tempo despendido entre os diferentes graus de dependência. / The quality of services offered by a health institution depends greatly on worker?s technical competence and interaction and communication abilities towards the client. Technical interventions performed by the nursing team require permanent evaluations of the risks involved. This observational and sectional study analyzed quality and time of execution of bathing and wound dressing procedures performed by the nursing team on hospitalized patients in a medical-surgical unit of a University Hospital in the State of Paraná, based on care dependency degrees of these clients. Target population was constituted by the following procedures: bathing, in three typologies (shower bath, shower bath with aid of a wheel chair and bed bath) and changing wound dressings on patients classified according to the nursing care dependency degree (I,II,III and IV). Convenience sampling was employed for both procedures. Data was collected through direct observation while using a check list instrument for documentation (verifying list). Procedure quality was analyzed based on the Positive Index (IP) for each item on the instrument and on the correct procedures score; execution time was evaluated in minutes. Quality of procedure execution was considered satisfactory when the IP median and the median correct procedure score was ? 70%. In this study 258 aspersion baths (42,6% degree I, 42,6& degree II and 14,8% degree III); 98 shower baths aided by wheel chair (12,5% degree I, 26,5% degree II, 54,1% degree III and 7,1% degree IV) and 46 bed baths (4,3% degree I, 37% degree III and 58,7% degree IV). The IP surpassed 70% only in the shower bath with aid of a wheel chair on patients classified as degree IV and on bed bath, degree I. The most frequently compromised items were related to orientation/communication/interaction with the patient, oral hygiene, bed disinfection, skin condition inspection and valuing patient complaints. Results obtained on the median of the correct procedures scores on the three bathing typologies, on the four dependency degrees also highlight low quality, because only in the shower bath with aid of a wheel chair, degree IV, and bed bath, degree I, 50% of the procedures reached 80% and 76,5% correct procedure scores, respectively. In the observed wound dressing procedures of the 168 patients, 33,9% were patients of degree I, 38,7% of degree II, 19,6% of degree III and 7,8% of degree IV. In all dependency degrees the recommended positivity index was reached (? 70%). Items such as adequate environment preparation, validity time frame checking, respect to aseptic principles and maintenance of procedure?s logical sequence, however, show low positivity. Medium scores were also superior to 70% in all dependency levels, indicating that the procedure meets a quality standard. It was not observed difference on time frame spent in the different dependency degrees in bathing and wound dressing procedures.

Image and video quality evaluation is very important. In applications involving signal transmission, the Reduced- or No-Reference quality metrics are generally more practical than the Full-Reference metrics. Digital watermarking based quality evaluation emerges as a potential Reduced- or No-Reference quality metric, which estimates signal quality by assessing the degradation of the embedded watermark. Since the watermark contains a small amount of information compared to the cover signal, performing accurate signal quality evaluation is a challenging task. Meanwhile, the watermarking process causes signal quality loss. To address these problems, in this thesis, a framework for image and video quality evaluation is proposed based on semi-fragile and adaptive watermarking. In this framework, adaptive watermark embedding strength is assigned by examining the signal quality degradation characteristics. The "Ideal Mapping Curve" is experimentally generated to relate watermark degradation to signal degradation so that the watermark degradation can be used to estimate the quality of distorted signals. With the proposed framework, a quantization based scheme is first implemented in DWT domain. In this scheme, the adaptive watermark embedding strengths are optimized by iteratively testing the image degradation characteristics under JPEG compression. This iterative process provides high accuracy for quality evaluation. However, it results in relatively high computational complexity. As an improvement, a tree structure based scheme is proposed to assign adaptive watermark embedding strengths by pre-estimating the signal degradation characteristics, which greatly improves the computational efficiency. The SPIHT tree structure and HVS masking are used to guide the watermark embedding, which greatly reduces the signal quality loss caused by watermark embedding. Experimental results show that the tree structure based scheme can evaluate image and video quality with high accuracy in terms of PSNR, wPSNR, JND, SSIM and VIF under JPEG compression, JPEG2000 compression, Gaussian low-pass filtering, Gaussian noise distortion, H.264 compression and packet loss related distortion.

Disertacijoje nagrinėjami GSM kanale vykstantys balso paketų praradimai. Nagrinėjama galimybė vertinti trumpų balso segmentų kokybę PESQ algoritmu. Pasiūlomas naujas, mažos trukmės balso signalų kokybės vertinimo PESQ algoritmu metodas. Pristatomas originalus balso kokybės vertinimo metodas mobiliojo ryšio sistemoms, paremtas balso kokybės klasėmis. / Thesis presents the research of voice frame losses in the GSM. The possibility of evaluation of quality of short duration signals using PESQ algorithm is analysed. New method for preparation of short duration signals for PESQ measurements is proposed. Original method based on quality classes for voice quality evaluation in mobile communications is presented.

Thesis presents the research of voice frame losses in the GSM. The possibility of evaluation of quality of short duration signals using PESQ algorithm is analysed. New method for preparation of short duration signals for PESQ measurements is proposed. Original method based on quality classes for voice quality evaluation in mobile communications is presented. / Disertacijoje nagrinėjami GSM kanale vykstantys balso paketų praradimai. Nagrinėjama galimybė vertinti trumpų balso segmentų kokybę PESQ algoritmu. Pasiūlomas naujas, mažos trukmės balso signalų kokybės vertinimo PESQ algoritmu metodas. Pristatomas originalus balso kokybės vertinimo metodas mobiliojo ryšio sistemoms, paremtas balso kokybės klasėmis.

Image and video quality evaluation is very important. In applications involving signal transmission, the Reduced- or No-Reference quality metrics are generally more practical than the Full-Reference metrics. Digital watermarking based quality evaluation emerges as a potential Reduced- or No-Reference quality metric, which estimates signal quality by assessing the degradation of the embedded watermark. Since the watermark contains a small amount of information compared to the cover signal, performing accurate signal quality evaluation is a challenging task. Meanwhile, the watermarking process causes signal quality loss. To address these problems, in this thesis, a framework for image and video quality evaluation is proposed based on semi-fragile and adaptive watermarking. In this framework, adaptive watermark embedding strength is assigned by examining the signal quality degradation characteristics. The "Ideal Mapping Curve" is experimentally generated to relate watermark degradation to signal degradation so that the watermark degradation can be used to estimate the quality of distorted signals. With the proposed framework, a quantization based scheme is first implemented in DWT domain. In this scheme, the adaptive watermark embedding strengths are optimized by iteratively testing the image degradation characteristics under JPEG compression. This iterative process provides high accuracy for quality evaluation. However, it results in relatively high computational complexity. As an improvement, a tree structure based scheme is proposed to assign adaptive watermark embedding strengths by pre-estimating the signal degradation characteristics, which greatly improves the computational efficiency. The SPIHT tree structure and HVS masking are used to guide the watermark embedding, which greatly reduces the signal quality loss caused by watermark embedding. Experimental results show that the tree structure based scheme can evaluate image and video quality with high accuracy in terms of PSNR, wPSNR, JND, SSIM and VIF under JPEG compression, JPEG2000 compression, Gaussian low-pass filtering, Gaussian noise distortion, H.264 compression and packet loss related distortion.