COMPUTING

Author: Pereira Loureiro Javier.
Year: 2003.
University: A CORUÑA [www.udc.es].
Place of defense: Facultad de Informática.
Place of preparation: Facultad de Informática.

Summary: In recent years, the introduction of the DICOM standard, the evolution of the Internet and continual improvements and advances in technology have led to an increase in the number of PACS installed in hospitals. These solutions still presents some problems: i) non - modular architecture ii) non-ergonomic interfaces iii) difficulty to track the patient's medical history as well as their integration in HIS, RIS iv) lack or absence of statistical analysis modules and processing v) not dispopnibilidad of APls vi) difficulties in the enforcement of regulations concerning the treatment of computerized health data All of these problems are caused mainly by the absence of a defined architecture and the use of proprietary solutions in those areas not defined by DICOM. This paper presents a new approach to the development of an information system for medical images that do not present the problems mentioned above. This is a new contribution to the evolutionary process of PACS, to define a model architectural priamidal based layers. We have identified four layers: network and acquisition; storage, processing and analysis, and system support for clinical decision making. The flow of information in the system comes contextualised space and time by their own data and analyzes to them. The base of the pyramid represents the totality of the data, while the top is the result of the processing and analysis of such data, providing information useful for clinical decision-making. Each layer represents an abstraction in the process of analyzing information and provides an interface to the layer immediately above. For each of these events will also study the standards that should be used. The interaction with other systems, both homogeneous and heterogeneous, is done via communication with other intra-and inter-hospital systems that support open standards. Security is an issue that involves all layers. From transmissions safe storage reliable and fault-tolerant to control access to data and the definition of security policies with different levels of privileges.

Author: VILLARET AUSELLE MATEU.
Year: 2003.
University: POLITÉCNICA DE CATALUÑA [www.upc.edu].
Place of defense: Sala C1. 001 Campus Nord.
Place of preparation: EDIFICI C6 Campus NORD.

Summary: In this thesis we present several results about Second-Order Unification. It is well known that the Second-Order Unification Problem is in general undecidable; the frontier between its decidable and undecidable subclasses is thin and it still has not been completely defined. Our purpose is to shed some light on the Second-Order Unification problem and study some of its variants. We have mainly focused our attention on Context Unification and Linear Second-Order Unification. Roughly speaking, these problems are variants of Second-Order Unification where second-order variables are required to be linear. Context Unification was defined more than ten years ago and its decidability has been an open question since then. Here we make relevant contributions to the study of this question. The first result that we present is a simplification on these problems thanks to "curryfication". We show that the Context Unification problem can be NP-reduced to the Context Unification problem where, apart from variables, just a single binary function symbol, and first-order constants, are used. We also show that a similar result also holds for Second-Order Unification. The main result of this thesis is the definition of a non-trivial sufficient and necessary condition on the unifiers, for the decidability of Context Unification. The condition is called rank-bound conjecture in order to enforce our belief about its truthness. It lies on a non-trivial measure of terms, the rank, and claims that, whenever an instance of the Context Unification problem is satisfiable, there exists a unifier with a rank not exceeding a certain bound depending on the size of the problem. Under the assumption of this conjecture, we give a reduction of the satisfiability problem for Context Unification to the (decidable) satisfiability problem of Word Unification with regular constraints. Finally, in the same spirit of the extension of Word Unification with regular constraints, we also study the natural extension of Context Unification by means of tree-regular constraints on variable instantiations. We contribute with two more results: - we establish a relationship between Linear Second-Order Unification and Context Unification. Mainly, we reduce Linear Second-Order Unification to Context Unification with tree-regular constraints, these constraints are used to avoid the capture of variables in this process. - We also establish a relationship between Context Unification and the Constraint Language for Lambda Structures. This last formalism is broadly used in the treatment of ambiguous sentences of natural language, and there is currently an effort to quantify its power, and define its computational nature. Relating this constraints language with the unification framework can help us to apply the theoretic results from one side to the other.

Author: JIMENEZ GONZALEZ DANIEL.
Year: 2003.
University: POLITÉCNICA DE CATALUÑA [www.upc.edu].
Place of defense: Dept. d'Arquitectura de Computadors. Modul C6-E106.
Place of preparation: EDIFICI D6 DESPATX 206 Campus NORD.

Summary: In this thesis we present and analyze algorithms and parallel sequencing exploiting the memory hierarchy of the computer used and / or reduce the communication of data. However, although the objectives of this thesis are the same as those of other jobs, how to achieve them is different. In this thesis we got those making the sorting algorithms are aware of the proposed computer architecture and features of the data that we manage. The data sets that we believe are sets that fit into main memory, but not in cache memory. The algorithms filed suit, at runtime, the characteristics of the data (duplicates, biased, etc.). To avoid yield losses depending on their characteristics. To do so, these algorithms do a partitioning of the data, using a technique called Mutant Reverse Sorting and what this thesis. Mutant Reverse Sorting dynamically adapts to the characteristics of the data and the computer. This technique analyzes a sample of the data set to order to select the fastest way to partition the data. This technique choose between Sorting and Counting Reverse Split depending on the data distribution. These techniques are also proposed in this thesis. The analysis of these techniques, along with the sorting algorithms presented, is done on computer-based IBM modules p630 processor Power4 and on computer SGI O2000 processors R10K. In the analysis for both computers are models of behavior are compared with actual executions. With all this, we got the sorting algorithms faster sequential and parallel to the characteristics of the data and the computers used. This is thanks to these algorithms are adapted to computers and characteristics of the data better than the rest of algorithms analyzed. So on the one hand, the sequential algorithm proposed SKC-Radix sort, get some performance improvements of over 25% compared with the best algorithms found in the literature. Moreover, the greater the number of duplicates or bias data, the greater the improvement achieved by the SKC-Radix sort. Moreover, the parallel algorithm presented PSKC-Radix sort, it can order up to 4 times more data than Load Balanced Radix sort in the same amount of time. Load Balanced Radix sort algorithm was the fastest in memory management and parallel to the dataset that command until the publication of our work.

Author: ESTEVES DE SOUSA JOSE MANUEL.
Year: 2003.
University: POLITÉCNICA DE CATALUÑA [www.upc.edu].
Place of defense: Facultat d 'Informàtica.
Place of preparation: EDIFICI C6 Campus NORD.

Summary: ERP is one the latest technologies that many organizations have undertaken. Typically, Enterprise Resource Planning (ERP) systems are software packages composed of several modules, such as human resources, sales, finance and production, providing cross-organizational integration of transaction-based data management throughout embedded business processes support. These software packages can be customized up to a certain limit to the specific needs of each organization. ERP was characterized as the most important development in the corporate use of technology in the 1990s. Unfortunately, many ERP projects have not been effective enough and hence have been unable to achieve all the results envisaged. As the cost of an ERP implementation project is very high, it is critical for an organization to make the project a success and start obtaining benefits out of it as fast as possible. But what is it that makes an ERP implementation project successful? To address this issue we propose the use of a Critical Success Factors (CSF) approach to manage ERP implementation projects. After an extensive literature review on ERP research and ERP implementation project studies, we have studied and have proposed results along the following issues: · The identification and definition of a comprehensive list of CSF. · The relevance of CSF along the typical ERP implementation phases. · The definition of Key Performance Indicators (KPI) for CSF. · The analysis of CSF management in some organizational contexts. A theoretical framework was developed in order to aid the process of answering the implied research questions. In order to accomplish the research aims of this research, we have proposed an interpretive research approach and a âmultimethod' research framework that combines various research methods, both quantitative and qualitative, with predominance of qualitative ones. · An annotated bibliography on ERP research. · A CSF unified model for ERP implementation projects. · A CSF relevance schema along the typical ERP implementation phases. · A new criticality indicator for Process Quality management (PQM) method. · A tentative set of KPI for some CSF and a systematic approach to develop the rest of KPI. · An ERP implementation model. · A CSF management analysis in two organizational contexts: a small and midsized enterprise and a public higher education institution. With regard to the case studies conducted, the first was a pilot case study of an ERP implementation in a Portuguese small and midsized enterprise. The second one was an in-depth case study of an ERP implementation in a big Spanish public higher education institution. The different organizational contexts provided valuable insights in CSF management as well as implications from the emergence of patterns of communality between both case studies. The research results evidence that: · Most of the problems that arise in ERP implementation projects are associated with the activities identified as CSF in this research, · The main concerns are organizational rather than technological. · The management of CSF is influenced by the context and, · When managers have taken into account the CSF identified, some of project problems have been avoided or their impact significantly reduced in ERP implementation projects, and the organization is more likely to use more effectively the ERP system after its implementation. · A CSF approach is also helpful to avoid problems on the long term since most of the CSF identified are strategic. It is hoped that future ERP research and ERP implementations can dra 8 w upon a 2a2 nd learn from this thesis.

Author: MIR RUBIO JUAN.
Year: 2004.
University: AUTÓNOMA DE BARCELONA [www.uab.es].
Place of defense: ESCOLA DE POSTGRAU.
Place of preparation: ESCUELA DE POSTGRADO.

Summary: The thesis contains the study and innovation made in the area of channels and anonymous in the field of security systems agents móviles.El focal point of these two lines of inquiry comes from the development of a channel using an anonymous system mobile agents. Internet users can communicate, to express their ideas and share their experiences, and to freely do need a fundamental property, privacy. The goal of a communications channel anonymous is to break the pattern of communication between users and achieve privacy in communications, property which is a right for users in Spain (see article 197.1 of the penal code). In this thesis introduces the use of the paradigm of mobile operators for the construction of a canal anonymous. Later, she makes a second proposal anonymous channel whose construction is simplified and finally provides a channel anonymous collaborative that suminsitra resistance to a group of t-adversarios. On the other hand the paradigm of mobile agents is ideal for those applications with a lot of data, where it is better to move the code to run in front of moving all of the data. An agent or mobile code móviel is a system formed by a code and the configuration of the environment, including data structures, the stack and control information as the meter program, and are expected to migrating between those servers offering an environment where run. Any system that has an execution environment for the actors was deneomina agency. To define the pattern of movements of the actors, what agencies visit and what tasks performed at each agency, a route is commonly used. Security in the mobile agents is a complex issue because of its generality and the large number of situations involving. The mobile operators must protect against the agencies where they are executed. Agencies can become malicious, itentando alter the code or data of the players for their own benefit. In this thesis presents work of protecting itineraries, ho area that had been tested until now. The author takes the first solution for each type of itinerary and later provides a comprehensive protocol protection.

Author: GONZÁLEZ SABATÉ JORDI.
Year: 2004.
University: AUTÓNOMA DE BARCELONA [www.uab.es].
Place of defense: FACULTAD DE INFORMÁTICA.
Place of preparation: ESCUELA DE POSTGRADO.

Author: PUJOL VILA ORIOL.
Year: 2004.
University: AUTÓNOMA DE BARCELONA [www.uab.es].
Place of defense: FACULTAD DE INFORMÁTICA.
Place of preparation: ESCUELA DE POSTGRADO.

Author: CASTELLÀ ROCA JORDI.
Year: 2004.
University: AUTÓNOMA DE BARCELONA [www.uab.es].
Place of defense: ESCOLA TÉCNICA SUPERIOR D'ENGINYERIA.
Place of preparation: ESCUELA DE POSTGRADO.

Summary: Networks of computers and especially the Internet have enabled some common activities such as buying or play can make remotely (e-shopping and e-gamgling). The game of poker through a computer network is known as mental poker. Keeping operational game at the same time ensuring the same standards of safety, fairness and auditing offered by casinos for the traditional poker is a complex problem. The most important aspects to consider when designing a protocol mental poker are as follows: the game functionality, security, the cost of computation and communication. The proposals in the literature typically focus solely on the first two points. This makes it difficult to know which of the proposals is more efficient providing the same functionality and security. The thesis begins with a formal analysis of the costs of the main proposals in the literature. The analysis is not limited to costs, but is studying the safety of each proposal, in fact, our study found a significant weakness in one of the protocols compared. The attack is presented in a separate chapter after comprehensive comparative analysis. The following three chapters of the thesis presents three new protocols to improve the proposals in the literature on different aspects. The first proposal belongs to the family of protocols without TTP and which do not preserve the confidentiality of the strategy of the strategy of the players. The second proposal is a protocol TTP without preserving the confidentiality of the strategy of the players. The protocol reduces the computational cost so that the players perform less mathematical operations. The third proposal introduced a new feature that typically offer no protocols in the literature, which is tolerance abandonment of the players. In other words, players can continue to play even if some players leave the game in the middle of the game.

Author: Villares Ferro Jesús.
Year: 2004.
University: A CORUÑA [www.udc.es].
Place of defense: Facultad de Informática.
Place of preparation: Facultad de Informática.

Summary: The relationship between Natural Language Processing and Information Retrieval is given by the variation lingüstica language, ie how the same concept can be expressed in different ways by changes in the expression, the biggest problem in the field of Recovery Information. The objective of this thesis is to develop the technology base for Natural Language Processing and study the feasibility of its application in systems Information Retrieval on documents in Spanish. While there are similar studies for other languages, with a clear command of English, Spanish has often been relegated to the background. Moreover, their greater complexity lingüstica does not allow an extrapolation immediate results for the English, demanding the implementation of specific experiments. However, we have had to cope with the lack of resources lingüsticos freely accessible to the Spanish. This has been due to restrict the complexity of the solutions proposed, focusing on the use of lexical information, obtaining easier. In addition, to minimize the computational cost for the implementation in practical environments, there has been extensive use of technology finite state. In this context has been developed, first, a pre-advanced base lingüstica for tokenización and segmenting texts. At flexivo, has been studied using the lemmatization in normalizing simple terms. At derivative, has developed a tool for automatic generation of families morphological-word sets linked derivativamente and who share the same root-for use in normalizing simple terms. A syntactic level, it has been tested using units as syntactic terms índicce complex. These units are generated by parsing using two surface analyzers: PATTERNS, based on patterns, and CASCADE. Cascades of translators based on finite. Finally, syntactic level, has been evaluated a new approach pseudo-sintáctica that uses a model based on similarities on the basis of distances.

Author: ARAGON LOPEZ RUBEN.
Year: 2004.
University: COMPLUTENSE DE MADRID [www.ucm.es].
Place of defense: CIENCIAS ECONOMICAS Y EMPRESARIALES.
Place of preparation: ECONOMICAS.

Summary: This dissertation focuses on the valuation of internet portals through real options for growth. We proceeded to conduct an analysis of the various existing methods. It also conducted a survey to leading European businesses with an emphasis on Terra Lycos, Tiscali and T. Online. These companies will also be analyzed taking into account the risk of failure which would explore various procedures and methods generally accepted. Finally, it raises new lines of research focusing on the valuation of intangible through real options.

Author: Caballero Muñoz-Reja Ismael.
Year: 2004.
University: CASTILLA-LA MANCHA [www.uclm.es].
Place of defense: Escuela Superior de Informática de Ciudad Real.
Place of preparation: Escuelas Superior de Informática de Ciudad Real.

Summary: Today, most organizations base their activity in information systems, the quality of which has been focused almost exclusively on traditional and software that operates them. However, they have neglected one hand data quality as a raw material in the production of information products, and secondly, the quality of their own information products. This requires the implementation of mechanisms to manage the quality of data and information to avoid a negative impact on the organizations. Many researchers, aware of these needs, have undertaken efforts to measure the quality of data and information and propose concrete models to solve specific problems, but the assessment shows that these jobs are usually dependent on a context of a domain , which subtracts them universality and applicability. There is therefore no universal framework that any organization can use to assess and improve the quality of data and information on a global and inclusive. This thesis entitled "Model for Evaluating and Improving the Quality Management of the Data and Information" discusses how organizations can improve the quality of data and information they manage through good management of . To that end, explores concepts such as quality of data and information, describes the parts or components of the information system to those concerns are guides on how to manage it and how to measure it to determine the negative impact that might cause on the organización; también se indica cómo mejorarla y también cómo dotar al Sistema de Información de cierta capacidad de gestión de la calidad de los datos para eliminar las causas de variabilidad y mejorar los procesos de producción y de gestión encargados del tratamiento de los datos. The previously declared intentions result in a framework for the assessment and improvement of the quality management of data and information. This proposal is structured on a key concept: the concept of Process Information Management (PGI) as a joint vision of production processes information and quality management data and information. The framework defines two key components: a Model for Quality Management Data and Information-based Levels of Maturity (CALDEA) structure that the objectives of quality of data and information that should reach a PGI, and Evaluation and Methodology Process Improvement Information Management (EVAMECAL) used to evaluate and improve the Process Information Management with reference to maturity levels of quality management model proposed. An organization can improve its level of quality of data and information identifying the PGI first, and then apply the framework of work beginning with those most critical. Combining the maturity levels of the various (and / or most important) Process Information Management in the Information System, it is possible to obtain an overall assessment of the level of maturity that the organization has on the data and information management and improve .

Author: CECHICH MILEN SUSANA ALEJANDRA.
Year: 2004.
University: CASTILLA-LA MANCHA [www.uclm.es].
Place of defense: ESCUELA SUPERIOR DE INFORMÁTICA.
Place of preparation: ESCUELA SUPERIOR DE INFORMATICA.

Summary: The filtering components COTS is to decide which alternatives should be selected for a better assessment. Once the alternatives have been evaluated assessment data are the basis for a decision-making process. This thesis presents a proposal that aims to help improve the way in which organizations do the filtering COTS components. The point of view is essential to address the problem by integrating technical and organizational aspects. The proposal is based on the concept of continuous improvement, as defined by Six-Sigma, which integrates the various aspects relevant to the filtrate. This thesis defines a process to support early detection of the adequacy of components with respect to a given specification. The elements of the process are: 1. A framework for the management of COTS components filtering based on Six-Sigma 2. A collection of framed procedures for the processes at every stage.

Author: REMOLAR QUINTANA INMACULADA.
Year: 2004.
University: JAUME I DE CASTELLON [www.uji.es].
Place of defense: E.S.TECNOLOGIA Y CIENCIAS EXPERIMENTALES.
Place of preparation: E.S. TECNOLOGIA Y CIENCIAS EXPERIMENTALES.

Summary: One of the most important issues in viewing interactive scenes is the external representation of the plant. The plants, especially trees, are extremely complex objects, that is, consisting of such a large number of polygons that practically prevents display real-time scenes in which they appear. While the hardware design is progressing rapidly, such amount of data far exceeds the capacity of existing graphics systems. This thesis addresses this problem and presents a new method for real-time display of trees and plants, which combines techniques based on geometry using techniques based on images. As for the geometric representation, presents a model that allows multirresolución vary with the level of detail that you see the foliage of the trees in a continuous manner. Moreover, this method is combined with visualization based on images, using mobile impostors. This will avail consistency between consecutive frames, inherent in the three-dimensional scenes, avoiding continually redraw the whole geometry of the scene, which allows for the interactive visualization of ecosystems, outdoor scenes with a large number of trees.

Author: ALVA OBESO MARIA ELENA.
Year: 2004.
University: OVIEDO [www.uniovi.es].
Place of defense: E.U.INGENIERIA TECNICA INFORMATICA.
Place of preparation: DEPARTAMENTO DE INFORMATICA.

Author: DÍEZ SÁNCHEZ ANA ISABEL.
Year: 2004.
University: VIGO [www.uvigo.es].
Place of defense: DEPARTAMENTO DE INFORMÁTICA.
Place of preparation: DEPARTAMENTO DE INFORMÁTICA.

Summary: The use of computers as a component of teaching and learning processes are remotely several decades ago and has gained a growing importance accentuated, if possible, by the globalization of communication and access to information furnished by the extension of the Internet and especially on the web [GONZ2003]. In many cases, the creation of specific software environments for teaching or learning is very low and teachers, having overcome the initial barrier to many it involves the use of computer, faced with the shortage of educational software products covering their needs at least the basic need to provide students a means of consultation at any time of our documentation and material to the subject. The philosophy of electronic books is becoming increasingly important in the area of education where the large volume of information used and the large audience that seeks access it creates a set of requirements that have been collected in different studies (ACM, IEEE , ITHET ,..). It is necessary to give answers that would allow for a proper use of these facilities and, at the same time, create a philosophy of using the inferfaces of metaphors appropriate for the user, and so on. To facilitate the appropriate use of information technologies. Therefore, this paper intends to focus on providing a different route to the different philosophies that have enabled the development of e-books, creating a new perspective for educators, based on different theories of teaching and learning. To this end, has developed a tool for generating advanced and studied by the different educational philosophies and serve as a system of synergistic work done by educators and educational institutions. TELL, which was developed under the paradigm of Object Orientation, allows the creation of electronic books and their subsequent display simulating the sensations that provides reading books on paper to guide its operation to the two key players in the process: the teacher and the pupils. Both blocks in the system and the teaching-learning process.

Author: DUCH BROWN AMALIA.
Year: 2004.
University: POLITÉCNICA DE CATALUÑA [www.upc.edu].
Place of defense: SALA DE JUNTAS FIB..
Place of preparation: EDIFICI C6 Campus NORD.

Author: MARTINEZ CARRERAS MARIA ANTONIA.
Year: 2004.
University: MURCIA [www.um.es].
Place of defense: FACULTAD DE INFORMATICA.
Place of preparation: FACULTAD DE INFORMATICA.

Summary: In recent years, is gaining importance on applications designed for collaboration between individuals or group of people. This boom can be seen in various areas of society, such as both the broad world of business or the educational field, possibly connected to a group of workers in the first case or a teacher with her students as well as group students in the second case. More specifically the use of information technology and communications (ICT) have become very important in the development of these applications. Thus emerges the field CSCW (Computer Supported Cooperative Work) with the aim of studying both the development of groupware applications as the impact they cause to society. The collaborative educational environments, as well as educational tools are taking great impact on our society, and takes place at present numerous research projects on its construction and its use in educational communities. The investigations conducted during these years have been based on educational environments, which have done a series of studies on the advantages and disadvantages that provide such environments, both as an architectural level tools. It is a fact that these environments offer a number of opportunities to enrich the student learning process. In developing an educational environment requires both pedagógos as programmers join their knowledge to obtain an environment that fits the needs of both proferores as students. In the investigations carried out have included two approaches in learning: the collaborative and cooperative. Through collaborative theories we analyzed the characteristics of a system CSCL well as a new field emerging within the latter, which is learning by discovery. To carry out the collaborative work is of vital importance to find an architecture that is based on collaboration and that is adaptable, reusable and allows incorporate a range of tools that enable users to engage with the environment. In this quest also claims that the tools developed can also be reused in different environments. There are currently CSCL educational systems that require the integration of new systems that enable synchronous collaboration, allowing such integration does not affect the architecture of the underlying environment. We face two arms research linked one of them is the quest for architecture to facilitate collaboration and the study of the features that should show the necessary tools to develop collaborative task and the other is the definition of a collaborative system easily integrated into any educational setting. Thus through projects EDUSI, PUPITRE-NET and "Evaluating Applications Tele-archivos, groupware and video conferencing, and its impact on Tele-Enseñanza within the Education of the Murcia region," the latter funded Established by the Seneca, we have been exploring and conducting a series of experiments with distance learning environment JLE called "Java Learning Environment", currently EDUSTANCE. Such an environment characterized by a cooperative learning, based on content, has served as a point of reference for the study and characterization of the tools that should be provided to facilitate the educational task both students and teachers. During the investigation conducted in the environment JLE, has studied the incorporation of tools synchronous conducting a series of studies to get the advantages and disadvantages of this system, which will serve for the collaborative design system. In the previous investigation to what constitutes our results in this paper, 8 inve 152f stigación has also conducted a study of architecture ANTS of which was obtained foundation for the propagation of events in collaboration synchronous. Through this platform has been studied both aspects for the inclusion of components and the creation of a generic architecture for connecting heterogeneous environments. These studies have provided a point of reference for the design of the collaborative platform which is one of our results. More specifically, the results of this research work has been partly funded by the European projects ITCOLE and COLAB. Through the work done in ITCOLE has developed a collaborative system that can be integrated into existing environments. As proof of concept of this system, the integration was conducted in two environments Web BSCL and FLE. The integration of this new collaborative system within each of the environments requires that the system meets the following objectives: "One of the main aspects is the communication between synchronous tools and underlying environment. Lifecycle tools synchronous communication and brings data transfer between the existing environment and new tools. Therefore the system should communicate underlying these statements to the environment for which it must provide appropriate methods. "Furthermore, the components of collaboration require a synchronous channel communication with high performance allowing the spread of the changes in the state in an area distributed. The spread of state imposes severe requirements on the channel of communication with the aim of securing communication one to one or one-to-many communication. "Another essential requirement for infrastructure is synchronous which is simple, compact, lightweight and easy to install. The main idea is to avoid the imposition of a surcharge in the underlying environment and therefore simplify the installation process for the end user. For this we propose a replicated architecture for synchronous components that are based on the channel or bus communication to propagate the state among the tools in place at the same meeting. "Another interesting aspect for dealing with the construction of tools is synchronous provide a model for the collection of events relevant to analyze later. As we shall see, to do so we rely on the use of channel events, so that a service listening events of interest and stored in the appropriate structures. Under the project COLAB has defined our architecture to support collaborative experimentation. Through this platform was built application COLAB for experimentation in the field of physics. So one of the goals of our research work is to provide a collaborative platform generic that can be applied to any field including collaborative learning by discovery. This platform that must meet the following requirements to architectural level: 1. The architecture must support asynchronous collaboration, in the sense that it must allow the saving of information dealt with at each meeting as well as the recovery of that information in subsequent interactions. 2. The architecture should provide a channel propagation of events for the exchange of information between different applications developed for synchronous collaboration. More specifically, using the model replicated in the collaborative programming tools so that events sent through the channel to the other partners update their views with the new information. 3. The information handled should be treated separate sessions, so that what takes place in the meeting of a particular group are not visible to other groups. 4. The architecture must be modular, such that allows for an easy way to introduce new tools and the inclusion of existing tools. In addition, the architecture must be built in such a way that allows separate part of the interface programming. 5. It must provide a clear and simple API to enable all members of the community management of this architecture in the design and integration of tools without having knowledge of all the technologies used in developing such a platform. 6. The platform should also include support for storing all the necessary information system and the resources to be used.

Author: SÁEZ MARTÍNEZ JUAN MANUEL.
Year: 2004.
University: ALICANTE [www.ua.es].
Place of defense: ESCUELA POLITÉCNICA SUPERIOR.
Place of preparation: ESCUELA POLITÉCNICA SUPERIOR.

Summary: In this paper we present an approach to the problem of Mapeado and Simultaneous Localization (SLAM), currently regarded as one of the basic pillars of mobile robotics. To do this, we turn to information appearance and structure 3D scene, obtained through a stereoscopic camera, which represents the single sensor system. This device offers observations semi-densas environment, with a noise not isotropic material. First, we present a rapid and robust algorithm to estimate the actions of the robot according to their observations, which allows us to dispense with a sensor odometria further. The algorithm performs estimating each share information with local, that is, using only the comments immediately before and after the course. To minimize the computational time, the calculation is made on a range of points of each observation (reduced observations). EI algorithm consists of three stages: initial pairing, where two relate to two points of the observations, using the appearance and information 3D; refinement, which are filtered falsely positive initial pairing, using the structure 3D observations and alignment, which believes the action taken on the basis of robust pairing obtained in Phase anterior.Bajo the assumption camera soil and straight, it is speeding up the process and minimize the ambigà ¼ of the same age. Removing this restriction, and some increase in time as in the computation of error of the estimate is achieved extend our previous solution to cover the entire camera movement (six degrees of freedom). Secondly, it proposes a technical correction overall. This technique uses all the information on the map to correct accumulated errors in the estimation phase of shares. This is an iterative scheme minimization of energy, which uses as a criterion for consistency entropy map. This measure allows find a map with a maximum overlap both local and global (long cycles), without the need for explicit distinctions between the two. Additionally, a term alignment based on entropy, applicable only on environments inside plano-paralelos (world Manhattan). This term allows rectify a map even in the absence of long cycles. Uniting the rectifiación global estimation of shares in the same scheme, we get our solution to the problem of Mapeado and Simultaneous Localization. The estimate of shares aorta approximate initial map, which takes as its starting point in the overall correction. The scheme also extends to SLAM case 6DOF, through an amendment of the role of energy in the overall phase correction. In both cases should be treated the scalability of the problem, although in the case 6DOF is critical. To that end, we propose a scheme incremental merger of observations, which is able to maintain the number of elements in the record within a manageable range. Moreover, in the case 6DOF raises a versíón anytime of the overall correction algorithm, which allows temporary need of a real-time system. Given the pattern of previous SLAM, proposes a technical exploration of the environment autonomously, inspired algorithms border. The sitemap is formed only two states (obstacles and goals), and the planning of the trajectory is done through a converted distance between them. This helps create a safe path, ie the route away from the obstacles that leads to the goal more quickly. The document is divided into two parts. The first contains the main body of the thesis, which consists of six chapters: Introduction and objectives, which raises the contents and obje 8 tive

Author: HERNÁNDEZ DEL OLMO FÉLIX.
Year: 2004.
University: NACIONAL DE EDUCACIÓN A DISTANCIA [www.uned.es].
Place of defense: E.T.S. INGENIERÍA INFORMÁTICA.
Place of preparation: E.T.S.I. INFORMÁTICA.

Summary: The paper presents a new framework that describes issues blurred, and often untreated, in the overall process of recommendation from around System Recomendador. This framework aims to resolve the questions what, when and how to be at least should be any website to provide information. If we take this as the ultimate goal to resolve any system Recomendador, introduced the framework will enable us to build concrete assessment methods, based on metric concrete to tell us how close we are to resolve those questions. All this, without sacrificing the generality required to be applied to any system Recomendador. So dedicate thesis to define this framework, and corroborate theoretical and experimental that it meets the foregoing. This new framework will be our main contribution to the evaluation of systems Recomendadores Adaptive.

Author: REVILLA DE LUCAS JESÚS.
Year: 2004.
University: NACIONAL DE EDUCACIÓN A DISTANCIA [www.uned.es].
Place of defense: E.T.S. INGENIERÍA INFORMÁTICA.
Place of preparation: E.T.S.I. INFORMÁTICA.

Summary: Based on the state studied art presents a series of extensions of the theories put forward: amplíaa synthesis algorithm controllers static through interactive dynamic drivers: extending the sufficient condition synthesis drivers H Infinite other behaviors, both for systems LTI ocmo PVL, applies the change nonlinear variable synthesis systems PVL.