Traduction statistique par recherche locale

La traduction statistique vise l’automatisation de la traduction par le biais de modèles statistiques. Dans ce travail, nous relevons un des grands défis du domaine : la recherche (Brown et al., 1993). Les systèmes de traduction statistique de référence, tel Moses (Koehn et al., 2007), effectuent généralement la recherche en explorant l’espace des préfixes par programmation dynamique, une solution coûteuse sur le plan computationnel pour ce problème potentiellement NP-complet (Knight, 1999). Nous postulons qu’une approche par recherche locale (Langlais et al., 2007) peut mener à des solutions tout aussi intéressantes en un temps et un espace mémoire beaucoup moins importants (Russell et Norvig, 2010). De plus, ce type de recherche facilite l’incorporation de modèles globaux qui nécessitent des traductions complètes et permet d’effectuer des modifications sur ces dernières de manière non-continue, deux tâches ardues lors de l’exploration de l’espace des préfixes. Nos expériences nous révèlent que la recherche locale en traduction statistique est une approche viable, s’inscrivant dans l’état de l’art.

Corps, pouvoir et douleur chronique : récits de vie de danseurs et de danseuses professionnels

La porte d’entrée méthodologique de cette recherche repose sur l’observation participante d’une patiente atteinte de «douleur chronique» adressée pour accompagnement thérapeutique par le Qi Gong. L’énoncé fréquent de situations de conflit avec la famille, avec l’employeur ou avec les médecins, m’a amenée à y considérer la similitude entre les symptômes énoncés, les usages du corps et les attitudes mentales en lien avec une représentation mécaniste et duelle du corps. Marcel Mauss a été le premier à désigner ce rapport entre les usages du corps et les paradigmes sociétaux et culturels, à l’œuvre. Dans une ère fortement marquée par la robotisation des entreprises et des services, la dématérialisation des documents et la tyrannie des nouvelles technologies, la «douleur chronique» vient donc questionner les usages du corps contemporain dans un monde du travail axé sur la performance et la productivité. Cette recherche a pour but d’interroger en quoi la «douleur chronique» serait un symptôme des apories contemporaines, dépassant ainsi largement le champ de la médecine et en quoi elle serait comme l’arme du faible, the weapon of the weak, en contexte d’oppression (Scott, 1985). Ultimement, elle questionne l’affirmation de Michel Foucault selon laquelle «nous avons tous du pouvoir dans le corps» (1987: 27). Le recueil de récits de vie auprès de professionnels de la danse, pressentis pour leur habileté à décrire leur ressenti, y déjoue une tendance à reléguer dans le domaine de la psyché ce qui ne fait pas la preuve de son évidence. La forte crédibilité de leur parole permet en outre de mieux documenter un phénomène complexe, entâché de beaucoup d’à priori et de révéler les dynamiques de pouvoir à l’œuvre au sein d’une profession confrontée aux limites corporelles, plus que tout autre, dans l’exercice de son art. Une approche généalogique de ces histoires de douleur et leur mise côte à côte permet d’identifier les «plis» que sont l’endurance et le rejet de toute médication face à la douleur et comment ils infléchissent son évolution. Une analyse plus approfondie des six d’entre elles qui s’en sont sorties y démontre que si le pouvoir s’exerce sur un corps perçu comme docile, il peut aussi s’inverser pour contribuer à la guérison, au prix d’une insurrection de savoirs enfouis, universels, mais «assujettis», et d’un changement de paradigme. Ainsi, une meilleure compréhension de la «douleur chronique» laisse entrevoir une possible réversibilité du phénomène et une réinsertion sociale, avec ou sans réorientation de carrière, à condition de repenser nos modes de représentation, de production, de relation au travail et aux usages du corps, le rapport à l’Autre et à soi.

JSreal : un réalisateur de texte pour la programmation web

Site web associé au mémoire: http://daou.st/JSreal

Les enjeux identitaires de l’humain dans le débat philosophique sur la robotique humanoïde et l’amélioration humaine

Article

I-SS: integrated supervision systems approach based on interactive components

Process supervision is the activity focused on monitoring the process operation in order to deduce conditions to maintain the normality including when faults are present Depending on the number/distribution/heterogeneity of variables, behaviour situations, sub-processes, etc. from processes, human operators and engineers do not easily manipulate the information. This leads to the necessity of automation of supervision activities. Nevertheless, the difficulty to deal with the information complicates the design and development of software applications. We present an approach called "integrated supervision systems". It proposes multiple supervisors coordination to supervise multiple sub-processes whose interactions permit one to supervise the global process

Upper limb robot mediated stroke therapy - GENTLE/s approach

Stroke is a leading cause of disability in particular affecting older people. Although the causes of stroke are well known and it is possible to reduce these risks, there is still a need to improve rehabilitation techniques. Early studies in the literature suggest that early intensive therapies can enhance a patient's recovery. According to physiotherapy literature, attention and motivation are key factors for motor relearning following stroke. Machine mediated therapy offers the potential to improve the outcome of stroke patients engaged on rehabilitation for upper limb motor impairment. Haptic interfaces are a particular group of robots that are attractive due to their ability to safely interact with humans. They can enhance traditional therapy tools, provide therapy "on demand" and can present accurate objective measurements of a patient's progression. Our recent studies suggest the use of tele-presence and VR-based systems can potentially motivate patients to exercise for longer periods of time. The creation of human-like trajectories is essential for retraining upper limb movements of people that have lost manipulation functions following stroke. By coupling models for human arm movement with haptic interfaces and VR technology it is possible to create a new class of robot mediated neuro rehabilitation tools. This paper provides an overview on different approaches to robot mediated therapy and describes a system based on haptics and virtual reality visualisation techniques, where particular emphasis is given to different control strategies for interaction derived from minimum jerk theory and the aid of virtual and mixed reality based exercises.

Fully automated software solution for protein quantitation by global metabolic labeling with stable isotopes

Metabolic stable isotope labeling is increasingly employed for accurate protein (and metabolite) quantitation using mass spectrometry (MS). It provides sample-specific isotopologues that can be used to facilitate comparative analysis of two or more samples. Stable Isotope Labeling by Amino acids in Cell culture (SILAC) has been used for almost a decade in proteomic research and analytical software solutions have been established that provide an easy and integrated workflow for elucidating sample abundance ratios for most MS data formats. While SILAC is a discrete labeling method using specific amino acids, global metabolic stable isotope labeling using isotopes such as (15)N labels the entire element content of the sample, i.e. for (15)N the entire peptide backbone in addition to all nitrogen-containing side chains. Although global metabolic labeling can deliver advantages with regard to isotope incorporation and costs, the requirements for data analysis are more demanding because, for instance for polypeptides, the mass difference introduced by the label depends on the amino acid composition. Consequently, there has been less progress on the automation of the data processing and mining steps for this type of protein quantitation. Here, we present a new integrated software solution for the quantitative analysis of protein expression in differential samples and show the benefits of high-resolution MS data in quantitative proteomic analyses.

Software evolution: a graph based model

A model based on graph isomorphisms is used to formalize software evolution. Step by step we narrow the search space by an informed selection of the attributes based on the current state-of-the-art in software engineering and generate a seed solution. We then traverse the resulting space using graph isomorphisms and other set operations over the vertex sets. The new solutions will preserve the desired attributes. The goal of defining an isomorphism based search mechanism is to construct predictors of evolution that can facilitate the automation of ’software factory’ paradigm. The model allows for automation via software tools implementing the concepts.

Software evolution: a graph based model

A model based on graph isomorphisms is used to formalize software evolution. Step by step we narrow the search space by an informed selection of the attributes based on the current state-of-the-art in software engineering and generate a seed solution. We then traverse the resulting space using graph isomorphisms and other set operations over the vertex sets. The new solutions will preserve the desired attributes. The goal of defining an isomorphism based search mechanism is to construct predictors of evolution that can facilitate the automation of ’software factory’ paradigm. The model allows for automation via software tools implementing the concepts.

Automated two-dimensional separation flow system with electrochemical preconcentration, stripping, capillary electrophoresis and contactless conductivity detection for trace metal ion analysis

This paper describes the automation of a fully electrochemical system for preconcentration, cleanup, separation and detection, comprising the hyphenation of a thin layer electrochemical flow cell with CE coupled with contactless conductivity detection (CE-C(4)D). Traces of heavy metal ions were extracted from the pulsed-flowing sample and accumulated on a glassy carbon working electrode by electroreduction for some minutes. Anodic stripping of the accumulated metals was synchronized with hydrodynamic injection into the capillary. The effect of the angle of the slant polished tip of the CE capillary and its orientation against the working electrode in the electrochemical preconcentration (EPC) flow cell and of the accumulation time were studied, aiming at maximum CE-C(4)D signal enhancement. After 6 min of EPC, enhancement factors close to 50 times were obtained for thallium, lead, cadmium and copper ions, and about 16 for zinc ions. Limits of detection below 25 nmol/L were estimated for all target analytes but zinc. A second separation dimension was added to the CE separation capabilities by staircase scanning of the potentiostatic deposition and/or stripping potentials of metal ions, as implemented with the EPC-CE-C(4)D flow system. A matrix exchange between the deposition and stripping steps, highly valuable for sample cleanup, can be straightforwardly programmed with the multi-pumping flow management system. The automated simultaneous determination of the traces of five accumulable heavy metals together with four non-accumulated alkaline and alkaline earth metals in a single run was demonstrated, to highlight the potentiality of the system.

Building a Learning Architecture : e-Learning and e-Services

With the rapid advancement of the webtechnology, more and more educationalresources, including software applications forteaching/learning methods, are available acrossthe web, which enables learners to access thelearning materials and use various ways oflearning at any time and any place. Moreover,various web-based teaching/learning approacheshave been developed during the last decade toenhance the capability of both educators andlearners. Particularly, researchers from bothcomputer science and education are workingtogether, collaboratively focusing ondevelopment of pedagogically enablingtechnologies which are believed to improve theinfrastructure of education systems andprocesses, including curriculum developmentmodels, teaching/learning methods, managementof educational resources, systematic organizationof communication and dissemination ofknowledge and skills required by and adapted tousers. Despite of its fast development, however,there are still great gaps between learningintentions, organization of supporting resources,management of educational structures,knowledge points to be learned and interknowledgepoint relationships such as prerequisites,assessment of learning outcomes, andtechnical and pedagogic approaches. Moreconcretely, the issues have been widelyaddressed in literature include a) availability andusefulness of resources, b) smooth integration ofvarious resources and their presentation, c)learners’ requirements and supposed learningoutcomes, d) automation of learning process interms of its schedule and interaction, and e)customization of the resources and agilemanagement of the learning services for deliveryas well as necessary human interferences.Considering these problems and bearing in mindthe advanced web technology of which weshould make full use, in this report we willaddress the following two aspects of systematicarchitecture of learning/teaching systems: 1)learning objects – a semantic description andorganization of learning resources using the webservice models and methods, and 2) learningservices discovery and learning goals match foreducational coordination and learning serviceplanning.

Real time video segmentation for recognising paint marks on bad wooden railway sleepers

Wooden railway sleeper inspections in Sweden are currently performed manually by a human operator; such inspections are based on visual analysis. Machine vision based approach has been done to emulate the visual abilities of human operator to enable automation of the process. Through this process bad sleepers are identified, and a spot is marked on it with specific color (blue in the current case) on the rail so that the maintenance operators are able to identify the spot and replace the sleeper. The motive of this thesis is to help the operators to identify those sleepers which are marked by color (spots), using an “Intelligent Vehicle” which is capable of running on the track. Capturing video while running on the track and segmenting the object of interest (spot) through this vehicle; we can automate this work and minimize the human intuitions. The video acquisition process depends on camera position and source light to obtain fine brightness in acquisition, we have tested 4 different types of combinations (camera position and source light) here to record the video and test the validity of proposed method. A sequence of real time rail frames are extracted from these videos and further processing (depending upon the data acquisition process) is done to identify the spots. After identification of spot each frame is divided in to 9 regions to know the particular region where the spot lies to avoid overlapping with noise, and so on. The proposed method will generate the information regarding in which region the spot lies, based on nine regions in each frame. From the generated results we have made some classification regarding data collection techniques, efficiency, time and speed. In this report, extensive experiments using image sequences from particular camera are reported and the experiments were done using intelligent vehicle as well as test vehicle and the results shows that we have achieved 95% success in identifying the spots when we use video as it is, in other method were we can skip some frames in pre-processing to increase the speed of video but the segmentation results we reduced to 85% and the time was very less compared to previous one. This shows the validity of proposed method in identification of spots lying on wooden railway sleepers where we can compromise between time and efficiency to get the desired result.

Provenance: The Bridge Between Experiments and Data

Current scientific applications are often structured as workflows and rely on workflow systems to compile abstract experiment designs into enactable workflows that utilise the best available resources. The automation of this step and of the workflow enactment, hides the details of how results have been produced. Knowing how compilation and enactment occurred allows results to be reconnected with the experiment design. We investigate how provenance helps scientists to connect their results with the actual execution that took place, their original experiment and its inputs and parameters.

Visões das gerências da auditoriainterna da Petrobras sobre a sua estratégia de atuação em facedas práticas de mercado e da nova modelagem organizacional da companhia

The purpose of this study is to identify the strategic vision of the Internal Audit Department of Petrel Brasileiro S.A. PETROBRAS, in comparison with the audit function's proposals and practices in competitive organizations and their reality. It also aims at the verification of the possible solutions, within Petrobras itself, so as to constantly add value to the business and to the shareholders. With this in mind, a research was carried out, contemplating the conceptual proposals and market practices related to auditing and to the current diagnosis and the organizational model of Petrobras, so as to choose elements for comparison and analysis of such vision. The results of this research pointed in the direction of questioning the organizational value of the internal audit action, concluding that it will only be possible to continuously attain such value by means of a permanent coordination with the organization's strategic level. This is especially true if the audit action participates effectively in the process of Corporate Governance, in defining the risks, the internal control system and the measurement of corporate performance, as related to the development of the strategic plan. However, any action along these lines is still heavily impacted and limited by several aspects of culture and relationship of the organizational power as well as by the beliefs of the organization and of the body of auditors. The involvement of the auditors as employees was also considered relevant in the auditing process, by means of the participation of the audited entities in self assessment2. This procedure is still not sufficiently guaranteed by successful experiments in major organizations, considering that there is not a clear demonstration of the effective benefits of adopting this practice, as ompared to the central control, seldom shared but strongly monitored by integrated information systems. Finally, this research points to the need to renew the concept of the formation and role of the auditors in modern competitive organizations, in the face of information technology and of automation of the instrument controls of the business. Therefore, one may conclude that the trend will be toward an action aiming at the revision of formal internal control matrixes, as they are established in such systems. On the other hand, the majority of audit human resources will be increasingly deployed to the evaluation of risk and control, as related to relevant events of a more abstract nature, as in the case of those connected with the uncontrollable factors of the external environment.