Deformation mechanisms in second-phase affected microstructures and their energy balance

Based on the relationship Zener parameter (Z=second-phase size/second-phase volume fraction) vs. calcite grain size (dg), second-phase controlled aggregates and microstructures that are weakly affected by second-phases are discriminated. The latter are characterized by large but constant grain sizes, high calcite grain boundary fractions and crystallographic preferred orientations (CPO), while calcite grain size and calcite grain boundary fraction decrease continuously and CPO weakens with decreasing Z in second-phase controlled microstructures. These observations suggest that second-phase controlled microstructures predominantly deform via granular flow because pinning of calcite grain boundaries reduces the efficiency of dynamic recrystallization favoring mass transfer processes and grain boundary sliding. In contrast, the balance of grain size reduction and growth by dynamic recrystallization maintains a steady state grain size in microstructures that are only weakly affected by second-phases promoting a predominance of dislocation creep. With increasing temperature, the relationship between Z and dg persists but the calcite grain size increases continuously. Based on microstructures, the energy of each modifying process is calculated and its relative contribution is compared with energies of the competing processes (surface energy, dragging energy, dynamic recrystallization energy). The steady state microstructures result from a temperature-dependent energy minimization procedure of the system.

SEQUENTIAL DECISIONS AND PREDICTIONS IN NATURAL LANGUAGE PROCESSING

Natural language processing has achieved great success in a wide range of ap- plications, producing both commercial language services and open-source language tools. However, most methods take a static or batch approach, assuming that the model has all information it needs and makes a one-time prediction. In this disser- tation, we study dynamic problems where the input comes in a sequence instead of all at once, and the output must be produced while the input is arriving. In these problems, predictions are often made based only on partial information. We see this dynamic setting in many real-time, interactive applications. These problems usually involve a trade-off between the amount of input received (cost) and the quality of the output prediction (accuracy). Therefore, the evaluation considers both objectives (e.g., plotting a Pareto curve). Our goal is to develop a formal understanding of sequential prediction and decision-making problems in natural language processing and to propose efficient solutions. Toward this end, we present meta-algorithms that take an existent batch model and produce a dynamic model to handle sequential inputs and outputs. Webuild our framework upon theories of Markov Decision Process (MDP), which allows learning to trade off competing objectives in a principled way. The main machine learning techniques we use are from imitation learning and reinforcement learning, and we advance current techniques to tackle problems arising in our settings. We evaluate our algorithm on a variety of applications, including dependency parsing, machine translation, and question answering. We show that our approach achieves a better cost-accuracy trade-off than the batch approach and heuristic-based decision- making approaches. We first propose a general framework for cost-sensitive prediction, where dif- ferent parts of the input come at different costs. We formulate a decision-making process that selects pieces of the input sequentially, and the selection is adaptive to each instance. Our approach is evaluated on both standard classification tasks and a structured prediction task (dependency parsing). We show that it achieves similar prediction quality to methods that use all input, while inducing a much smaller cost. Next, we extend the framework to problems where the input is revealed incremen- tally in a fixed order. We study two applications: simultaneous machine translation and quiz bowl (incremental text classification). We discuss challenges in this set- ting and show that adding domain knowledge eases the decision-making problem. A central theme throughout the chapters is an MDP formulation of a challenging problem with sequential input/output and trade-off decisions, accompanied by a learning algorithm that solves the MDP.

Clinical Studies on conformal radiotherapy combined with epidermal growth factor receptor-tyrosine kinase inhibitor in second-line treatment of non-small cell lung cancer

Purpose: To study the effect of conformal radiotherapy combined with epidermal growth factor receptor-tyrosine kinase inhibitor (EGFR-TKI) in the second-line treatment of non-small cell lung cancer (NSCLC). Methods: A total of 316 patients attending Shanghai Pulmonary Hospital affiliated to Tongji University, were divided into two groups: 106 patients were treated with conformal radiotherapy combined with EGFR-TKI (gefitinib, 250 mg/day; or erlotinib, 150 mg/day), while 210 patients were treated with EGFRTKI alone. Some factors, including adverse reactions (AR), disease control rate (DCR), progression-free survival (PFS), overall survival (OS), and one-year and two-year survival rate, were evaluated. Results: No obvious difference was observed in AR between the two groups (p > 0.05). In the combination therapy group, complete response (CR) was 5 cases, partial response (PR) 43 cases, and stable disease (SD) 47 cases, progressive disease (PD) was 11 cases, response rate (RR) was 45.3 %, and DCR 89.6 %. Median PFS in the combination therapy group and targeted therapy group was 6.5 and 5.0 months, respectively. On the other hand, median OS in the combination therapy group and targeted group was 14.1 and 12.6 months, respectively. One-year survival rate of the combination therapy group and EGFR-TKI group was 60.3 and 50.0 %, respectively, while the two-year survival rate was 26.3 and 19.0 %, respectively. Conclusion: Conformal radiotherapy combined with EGFR-TKI can be used as an effective second-line treatment for NSCLC.

Reduced vowel production in American English among Spanish-English bilinguals

Prominent views in second language acquisition suggest that the age of L2 learning is inversely correlated with native-like pronunciation (Scovel, 1988; Birdsong, 1999). The relationship has been defined in terms of the Critical Period Hypothesis, whereby various aspects of neural cognition simultaneously occur near the onset of puberty, thus inhibiting L2 phonological acquisition. The current study tests this claim of a chronological decline in pronunciation aptitude through the examination of a key trait of American English – reduced vowels, or “schwas.” Groups of monolingual, early bilingual, and late bilingual participants were directly compared across a variety of environments phonologically conditioned for vowel reduction. Results indicate that late bilinguals have greater degrees of difficulty in producing schwas, as expected. Results further suggest that the degree of differentiation between schwa is larger than previously identified and that these subtle differences may likely be a contributive factor to the perception of a foreign accent in bilingual speakers.

Seniors in Second Life

Relatório de estágio apresentado à Escola Superior de Educação do Instituto Politécnico de Santarém para a obtenção do grau de mestre em Educação e Comunicação Multimédia

Role of the Tryptophan Residues in the Specific interaction of the Sea Anemone Stichodacty la helianthus’s Actinoporin Sticholysin II with Biological Membranes

Actinoporins are pore-forming toxins from sea anemones. Upon interaction with sphingomyelin-containing bilayers, they become integral oligomeric membrane structures that form a pore. Sticholysin II from Stichodactyla helianthus contains five tryptophans located at strategic positions; its role has now been studied using different mutants. Results show that W43 and W115 play a eterminant role in maintaining the high thermostability of the protein, while W146 provides specific interactions for protomer−protomer assembly. W110 and W114 sustain the hydrophobic effect, which is one of the major driving forces for membrane binding in the presence of Chol. However, in its absence, additional interactions with sphingomyelin are required. These conclusions were confirmed with two sphingomyelin analogues, one of which had impaired hydrogen bonding properties. The results obtained support actinoporins’ Trp residues playing a major role in membrane recognition and binding, but their residues have an only minor influence on the diffusion and oligomerization steps needed to assemble a functional pore.

Using Research Methods in Human Computer Interaction to Design Technology for Resilience

Research in human computer interaction (HCI) covers both technological and human behavioural concerns. As a consequence, the contributions made in HCI research tend to be aware to either engineering or the social sciences. In HCI the purpose of practical research contributions is to reveal unknown insights about human behaviour and its relationship to technology. Practical research methods normally used in HCI include formal experiments, field experiments, field studies, interviews, focus groups, surveys, usability tests, case studies, diary studies, ethnography, contextual inquiry, experience sampling, and automated data collection. In this paper, we report on our experience using the evaluation methods focus groups, surveys and interviews and how we adopted these methods to develop artefacts: either interface’s design or information and technological systems. Four projects are examples of the different methods application to gather information about user’s wants, habits, practices, concerns and preferences. The goal was to build an understanding of the attitudes and satisfaction of the people who might interact with a technological artefact or information system. Conversely, we intended to design for information systems and technological applications, to promote resilience in organisations (a set of routines that allow to recover from obstacles) and user’s experiences. Organisations can here also be viewed within a system approach, which means that the system perturbations even failures could be characterized and improved. The term resilience has been applied to everything from the real estate, to the economy, sports, events, business, psychology, and more. In this study, we highlight that resilience is also made up of a number of different skills and abilities (self-awareness, creating meaning from other experiences, self-efficacy, optimism, and building strong relationships) that are a few foundational ingredients, which people should use along with the process of enhancing an organisation’s resilience. Resilience enhances knowledge of resources available to people confronting existing problems.

Seniors in Second Life

Relatório de estágio apresentado à Escola Superior de Educação do Instituto Politécnico de Santarém para a obtenção do grau de mestre em Educação e Comunicação Multimédia

Towards Unstructured Knowledge Integration in Natural Language Processing

In the last decades, Artificial Intelligence has witnessed multiple breakthroughs in deep learning. In particular, purely data-driven approaches have opened to a wide variety of successful applications due to the large availability of data. Nonetheless, the integration of prior knowledge is still required to compensate for specific issues like lack of generalization from limited data, fairness, robustness, and biases. In this thesis, we analyze the methodology of integrating knowledge into deep learning models in the field of Natural Language Processing (NLP). We start by remarking on the importance of knowledge integration. We highlight the possible shortcomings of these approaches and investigate the implications of integrating unstructured textual knowledge. We introduce Unstructured Knowledge Integration (UKI) as the process of integrating unstructured knowledge into machine learning models. We discuss UKI in the field of NLP, where knowledge is represented in a natural language format. We identify UKI as a complex process comprised of multiple sub-processes, different knowledge types, and knowledge integration properties to guarantee. We remark on the challenges of integrating unstructured textual knowledge and bridge connections with well-known research areas in NLP. We provide a unified vision of structured knowledge extraction (KE) and UKI by identifying KE as a sub-process of UKI. We investigate some challenging scenarios where structured knowledge is not a feasible prior assumption and formulate each task from the point of view of UKI. We adopt simple yet effective neural architectures and discuss the challenges of such an approach. Finally, we identify KE as a form of symbolic representation. From this perspective, we remark on the need of defining sophisticated UKI processes to verify the validity of knowledge integration. To this end, we foresee frameworks capable of combining symbolic and sub-symbolic representations for learning as a solution.

L' Ensenyament per tasques: programació d'una assignatura de traducció de l'alemany a l'espanyol

In this work we present a proposal for a course in translation from German into Spanish following the task based approach as known in second language acquisition. The aim is to improve the translation competence of translation students. We depart from the hypothesis that some students select inapropiate translation strategies when faced with certain translation problems leading them to translation errors. In order to avoid these translation errors originated by wrong application of such strategies we propose a didactic method which helps to prevent them by a) raising awareness of the different subcompetences required while translating, b) improving the ability to identify translation problems and relate them to the different subcompetences and c) enhancing the use of the most adequate strategy according to the characteristics of each problem. With regard to translation and how translation competence is acquired our work follows the communicative approach to translation theory as defended among others by Hatim & Mason (1990), Lörscher (1992) and Kiraly (1995), where translation is seen as a communicative activity which can be analized from a psycholinguistic perspective. In this sense we give operative definitions for what we understand by “translation problem”, “translation strategy”, “translation error”, “translation competence” and “translation”. Our approach to didactics adapts recent developments in Second Language Teaching within the communicative paradigm as is the task based approach by Nunan (1989) acquisition to translation. Fitting the recquirements of this pedagogic approach we present a planning for a translation course which is compatible with present translation studies.