Improving Collaborative Filtering Based Recommender Systems Using Pareto Dominance

Los sistemas de recomendación son un tipo de solución al problema de sobrecarga de información que sufren los usuarios de los sitios web en los que se pueden votar ciertos artículos. El sistema de recomendación de filtrado colaborativo es considerado como el método con más éxito debido a que sus recomendaciones se hacen basándose en los votos de usuarios similares a un usuario activo. Sin embargo, el método de filtrado de colaboración tradicional selecciona usuarios insuficientemente representativos como vecinos de cada usuario activo. Esto significa que las recomendaciones hechas a posteriori no son lo suficientemente precisas. El método propuesto en esta tesis realiza un pre-filtrado del proceso, mediante el uso de dominancia de Pareto, que elimina los usuarios menos representativos del proceso de selección k-vecino y mantiene los más prometedores. Los resultados de los experimentos realizados en MovieLens y Netflix muestran una mejora significativa en todas las medidas de calidad estudiadas en la aplicación del método propuesto. ABSTRACTRecommender systems are a type of solution to the information overload problem suffered by users of websites on which they can rate certain items. The Collaborative Filtering Recommender System is considered to be the most successful approach as it make its recommendations based on votes of users similar to an active user. Nevertheless, the traditional collaborative filtering method selects insufficiently representative users as neighbors of each active user. This means that the recommendations made a posteriori are not precise enough. The method proposed in this thesis performs a pre-filtering process, by using Pareto dominance, which eliminates the less representative users from the k-neighbor selection process and keeps the most promising ones. The results from the experiments performed on Movielens and Netflix show a significant improvement in all the quality measures studied on applying the proposed method.

Using multimedia and peer assessment to promote collaborative e-learning

Collaborative e-learning is increasingly appealing as a pedagogical approach that can positively affect student learning. We propose a didactical model that integrates multimedia with collaborative tools and peer assessment to foster collaborative e-learning. In this paper, we explain it and present the results of its application to the “International Seminars on Materials Science” online course. The proposed didactical model consists of five educational activities. In the first three, students review the multimedia resources proposed by the teacher in collaboration with their classmates. Then, in the last two activities, they create their own multimedia resources and assess those created by their classmates. These activities foster communication and collaboration among students and their ability to use and create multimedia resources. Our purpose is to encourage the creativity, motivation, and dynamism of the learning process for both teachers and students.

RECLAMO: virtual and collaborative honeynets based on trust management and autonomous systems applied to intrusion management

Security intrusions in large systems is a problem due to its lack of scalability with the current IDS-based approaches. This paper describes the RECLAMO project, where an architecture for an Automated Intrusion Response System (AIRS) is being proposed. This system will infer the most appropriate response for a given attack, taking into account the attack type, context information, and the trust and reputation of the reporting IDSs. RECLAMO is proposing a novel approach: diverting the attack to a specific honeynet that has been dynamically built based on the attack information. Among all components forming the RECLAMO's architecture, this paper is mainly focused on defining a trust and reputation management model, essential to recognize if IDSs are exposing an honest behavior in order to accept their alerts as true. Experimental results confirm that our model helps to encourage or discourage the launch of the automatic reaction process.

Outline of Europe-Japan collaborative research on concentrator photovoltaics

The Europe-Japan Collaborative Research Project on Concentrator Photovoltaics (CPV) has been initiated under support by the EC (European Commission) and NEDO (New Energy and Industrial Technology Development Organization) since June 2011. This is project (NGCPV Project; a New Generation of Concentrator PhotoVoltaic cells, modules and systems) is aiming to accelerate the move to very high efficiency and lower cost CPV technologies and to enhance widespread deployment of CPV systems. 7 organizations such as UPM, FhG-ISE Imperial College, BSQ, CEA-INES, ENEA, and PSE in Europe and 9 organizations such as TTI, Univ. Tokyo, AIST, Sharp Co. Daido Steel Co., Kobe Univ., Miyazaki Univ., Asahi Kasei Co., and Takano Co. participate in this project. The targets of this project are 1) to develop world-record efficiency CPV cells of more than 45%, 2) to develop world-record efficiency CPV modules of 35%, 3) to establish standard measurements of CPV cells and modules, 4) to install 50kW CPV system in Spain, to carry out field test of CPV system and to manage power generation of CPV systems, and 5) to develop high-efficiency and low-cost new materials and structure cells such as III-V-N, III-V-on-Si tandem, quantum dots and wells. This paper presents outline of this project and most recent results such as world record efficiency (37.9% under 1-sun) cell and high-efficiency (43.5% under 240-306 suns) concentrator cell with inverted epitaxial grown InGaP/GaAs/InGaAs 3-junction solar cells.

Collaborative Engineering: an Airbus case study

This document introduces the main concepts of Collaborative Engineering as a new methodology, procedures and tools to design and develop an aircraft, as Airbus Military is implementing. Airbus designs and industrializes aircrafts under Concurrent Engineering techniques since decades with success. The introduction of new PLM methodologies, procedures and tools, mainly in the industrialization areas, and the need to reduce time-to-market conducted Airbus Military to push the engineering teams to do things in a different way. Traditional Engineering works sequentially, Concurrent Engineering basically overlaps tasks between teams using maturity states and taking assuming risks. Collaborative Engineering promotes a single team to develop product, processes and resources from the conceptual phase to the start of the serial production. The deliverable of the team is an iDMU (industrial DMU), a complete definition and verification of the virtual manufacturing of the product.

Collaborative Engineering Paradigm Applied to the Aerospace Industry

Airbus designs and industrializes aircrafts using Concurrent Engineering techniques since decades. The introduction of new PLM methods, procedures and tools, and the need to reduce time-to-market, led Airbus Military to pursue new working methods. Traditional Engineering works sequentially. Concurrent Engineering basically overlaps tasks between teams. Collaborative Engineering promotes teamwork to develop product, processes and resources from the conceptual phase to the start of the serial production. The CALIPSO-neo pilot project was launched to support the industrialization process of a medium size aerostructure. The aim is to implement the industrial Digital Mock-Up (iDMU) concept and its exploitation to create shop floor documentation. In a framework of a collaborative engineering strategy, the project is part of the efforts to deploy Digital Manufacturing as a key technology for the industrialization of aircraft assembly lines. This paper presents the context, the conceptual approach and the methodology adopted.

Hierarchical graph maps for visualization of collaborative recommender systems

In this paper we provide a method that allows the visualization of similarity relationships present between items of collaborative filtering recommender systems, as well as the relative importance of each of these. The objective is to offer visual representations of the recommender system?s set of items and of their relationships; these graphs show us where the most representative information can be found and which items are rated in a more similar way by the recommender system?s community of users. The visual representations achieved take the shape of phylogenetic trees, displaying the numerical similarity and the reliability between each pair of items considered to be similar. As a case study we provide the results obtained using the public database Movielens 1M, which contains 3900 movies.

Collaborative Evolution Strategies on Evolvable Hardware Networked Elements

Dynamic and Partial Reconfiguration allows systems to change some parts of their hardware at run time. This feature favours the inclusion of evolutionary strategies to provide optimised solutions to the same problem so that they can be mixed and compared in a way that only the best ones prevail. At the same time, distributed intelligence permits systems to work in a collaborative way to jointly improve their global capabilities. This work presents a combination of both approaches where hardware evolution is performed both at local and network level in order to improve an image filter application in terms of performance, robustness and providing the capacity of avoiding local minimums, which is the main drawback of some evolutionary approaches.

El enfoque integrado en los programas de regeneración urbana de barrios desfavorecidos: una visión relacional

Actualmente se reconoce de forma generalizada que la problemática de los barrios más desfavorecidos presenta un trasfondo estructural de magnitud multidimensional que implica a múltiples actores, que requiere respuestas de carácter multiescalar, multidimensional y multiagente y un alto nivel de cooperación institucional y de coordinación entre la diversidad de actores, que pone en cuestión los modelos de gobernanza y gestión urbana tradicionales. Frente a los retos e implicaciones que plantea la noción de espacio como construcción social y los problemas de coordinación detectados que limitan el alcance de la regeneración urbana integrada, en esta investigación se pregunta por los principales elementos que pueden facilitar la dinamización y coordinación de los procesos y se plantea que es preciso aprovechar el propio proceso participativo para reforzar el papel de estos elementos y generar mejores condiciones para la colaboración en un proceso de aprendizaje práctico de formas de análisis e intervención integrada y gestión relacional, que supere los niveles de información y consulta de la práctica participativa habitual. Se centra por tanto la atención en los factores que facilitan la dinamización y coordinación de los procesos (elementos dinamizadores y articuladores), en el nivel de participación social y cooperación institucional en los distintos momentos de la planificación de la estrategia entendida como un proceso continuo de reflexión-acción-reflexión y en aquellas condiciones del modelo de participación que pueden propiciar relaciones estratégicas y mejorar la capacidad de interacción y acción del sistema de actores y relaciones en el proceso de planificación y gestión, que lleva a cuestionar en qué medida los soportes (foros de participación y nodos y ejes de actividad y relación) y el tipo de técnicas (técnicas de dinamización y técnicas de identificación y activación de oportunidades) de los modelos de participación de experiencias concretas pueden fomentar formas de interacción que mejoren la comunicación y la movilización de recursos. Para analizar esta situación en la práctica, se desarrolla un marco conceptual, metodológico y operativo a partir de la contribución de conceptos, teorías y modelos basados en una visión relacional y multidimensional de las dinámicas de desarrollo local integrado. Esta metodología se valida en cuatro casos de referencia a partir de la información de documentos oficiales y de las visiones de la diversidad de actores implicados recogidas a través de una primera consulta a expertos (13) y entrevistas semiestructuradas a representantes de la diversidad de actores implicados (75), que se ha procesado por medio de tres herramientas conceptuales de análisis complementarias desarrolladas a lo largo de la investigación apoyadas en la perspectiva relacional de la reinterpretación actual del espacio y las escalas: el ‘Sistema de acción local’, el ‘Diagrama de momentos’ del proceso de planificación y gestión, y la ‘Matriz de caracterización’ del modelo de participación. La evaluación y la comparativa de los casos según el método de análisis evidencian la importancia del papel ejercido por los elementos articuladores (actores y organismos de enlace, espacios de encuentro y acción, y los significados y visiones compartidas) y dinamizadores (el capital social y las capacidades institucionales propiciadas en redes de colaboración) a la hora de facilitar la colaboración entre las distintas instituciones responsables, reforzar las relaciones entre actores, grupos y entidades del barrio, y mejorar los problemas de convivencia vecinal, y en definitiva, de mejorar las posibilidades de los diferentes actores urbanos de implicarse de manera activa. Además, las fórmulas alternativas que se han puesto en marcha en los distintos casos frente a las dificultades que presenta la administración a la hora de trabajar por proyectos y el resto de actores de colaborar de forma constructiva, han supuesto un importante proceso de aprendizaje y práctica en formas de participación y colaboración. ABSTRACT It is widely accepted that the current situation of multi-deprived neighborhoods has a structural and a multidimensional basis involving multiple actors that calls for multiscale, multidimensional and multiagent solutions and requires a high level of institutional cooperation and a great deal of coordination between different actors. This calls into question traditional governance and management models. In order to cope with the challenges and implications of the idea of the social construction of space and these coordination requirements that compromise the scope of integrated urban regeneration strategies, this research focuses on the main elements and factors that facilitate the mobilization and coordination of the regeneration processes, and recognizes the importance of participation processes facilitating better social, communication and collaboration skills in a practical learning process of integrated analysis and urban regeneration interventions and relational governance mechanisms going beyond mere information and consultation levels. Thus, the study focuses on the principal factors which facilitate the mobilization and coordination of integrated area regeneration, the level of social participation and institutional cooperation at different planning moments in a continuous process of reflection-action-reflection, and the conditions under which the participation process may promote strategic relationships and improve the performance of the system of actors and relationships in the planning and management process. This questions to what extent the supports (participation arenas and activity nodes and axis) and the type of techniques (participation encouragement and opportunities identification and activation techniques) considered in participation models had led to a level of interaction that has improved communication and resource mobilization capabilities. In order to analyze this situation in practice, a conceptual, methodological and operational framework is developed, which considers the contributions of the review of concepts, theories and models based on a relational and multidimensional view of local development dynamics. This methodology is validated in four representative case studies in view of official documents and the contributions of diverse stakeholders gathered through a first consultation to experts (13) and semi-structured interviews to representatives of the diversity of actors involved (75). This data has been processed by three complementary analytical tools developed during the research, based on the relational perspective of the current reinterpretation of space and scales: the 'Local Action System', the 'Diagram of planning moments', and the 'Characterization matrix of the participation model'. The comparative evaluation of the four case studies carried out using the analytical method developed during the project, shows the importance of the availability and the role of linkage elements (actors and organizations, collective spaces and shared visions) and drivers (social capital and institutional capacities generated through collaborative networks) in facilitating the collaboration among all responsible institutions, in promoting and strengthening relations between actors, groups and neighborhood organizations, and in improving the conviviality and social cohesion conditions, and ultimately in fostering the possibilities of diversity of urban actors to engage and participate actively in the transformation process, increasing accordingly the options for change. In addition, despite all the difficulties, these practices have implemented alternative formulas addressing the limitations of the administration in working by projects and not by competences and of the stakeholders in collaborating in a constructive way. The development and implementation of these participatory formulas has been an important learning process.

Nuevo paradigma para decisiones colectivas en entornos no competitivos

La tesis estudia uno de los aspectos más importantes de la gestión de la sociedad de la información: conocer la manera en que una persona valora cualquier situación. Esto es importante para el individuo que realiza la valoración y para el entorno con el que se relaciona. La valoración es el resultado de la comparación: se asignan los mismos valores a alternativas similares y mayores valores a alternativas mejor consideradas en el proceso de comparación. Los patrones que guían al individuo a la hora de hacer la comparación se derivan de sus preferencias individuales (es decir, de sus opiniones). En la tesis se presentan varios procedimientos para establecer las relaciones de preferencia entre alternativas de una persona. La valoración progresa hasta obtener una representación numérica de sus preferencias. Cuando la representación de preferencias es homogénea permite, además, contrastar las preferencias personales con las del resto de evaluadores, lo que favorece la evaluación de políticas, la transferencia de información entre diferentes individuos y el diseño de la alternativa que mejor se adapte a las preferencias identificadas. Al mismo tiempo, con esta información se pueden construir comunidades de personas con los mismos sistemas de preferencias ante una cuestión concreta. La tesis muestra un caso de aplicación de esta metodología: optimización de las políticas laborales en un mercado real. Para apoyar a los demandantes de empleo (en su iniciación o reincorporación al mundo laboral o en el cambio de su actividad) es necesario conocer sus preferencias respecto a las ocupaciones que están dispuestos a desempeñar. Además, para que la intermediación laboral sea efectiva, las ocupaciones buscadas deben de ser ofrecidas por el mercado de trabajo y el demandante debe reunir las condiciones para acceder a esas ocupaciones. El siguiente desarrollo de estos modelos nos lleva a los procedimientos utilizados para transformar múltiples preferencias en una decisión agregada y que consideran tanto la opinión de cada uno de los individuos que participan en la decisión como las interacciones sociales, todo ello dirigido a generar una solución que se ajuste lo mejor posible al punto de vista de toda la población. Las decisiones con múltiples participantes inciden, principalmente, en: el aumento del alcance para incorporar a personas que tradicionalmente no han sido consideradas en las tomas de decisiones, la agregación de las preferencias de los múltiples participantes en las tomas de decisiones colectivas (mediante votación, utilizando aplicaciones desarrolladas para la Web2.0 y a través de comparaciones interpersonales de utilidad) y, finalmente, la auto-organización para permitir que interaccionen entre si los participantes en la valoración, de forma que hagan que el resultado final sea mejor que la mera agregación de opiniones individuales. La tesis analiza los sistemas de e-democracia o herramientas para su implantación que tienen más más utilización en la actualidad o son más avanzados. Están muy relacionados con la web 2.0 y su implantación está suponiendo una evolución de la democracia actual. También se estudian aplicaciones de software de Colaboración en la toma de decisiones (Collaborative decision-making (CDM)) que ayudan a dar sentido y significado a los datos. Pretenden coordinar las funciones y características necesarias para llegar a decisiones colectivas oportunas, lo que permite a todos los interesados participar en el proceso. La tesis finaliza con la presentación de un nuevo modelo o paradigma en la toma de decisiones con múltiples participantes. El desarrollo se apoya en el cálculo de las funciones de utilidad empática. Busca la colaboración entre los individuos para que la toma de decisiones sea más efectiva, además pretende aumentar el número de personas implicadas. Estudia las interacciones y la retroalimentación entre ciudadanos, ya que la influencia de unos ciudadanos en los otros es fundamental para los procesos de toma de decisiones colectivas y de e-democracia. También incluye métodos para detectar cuando se ha estancado el proceso y se debe interrumpir. Este modelo se aplica a la consulta de los ciudadanos de un municipio sobre la oportunidad de implantar carriles-bici y las características que deben tomar. Se simula la votación e interacción entre los votantes. ABSTRACT The thesis examines one of the most important aspects of the management of the information society: get to know how a person values every situation. This is important for the individual performing the assessment and for the environment with which he interacts. The assessment is a result of the comparison: identical values are allocated to similar alternatives and higher values are assigned to those alternatives that are more favorably considered in the comparison process. Patterns that guide the individual in making the comparison are derived from his individual preferences (ie, his opinions). In the thesis several procedures to establish preference relations between alternatives a person are presented. The assessment progresses to obtain a numerical representation of his preferences. When the representation of preferences is homogeneous, it also allows the personal preferences of each individual to be compared with those of other evaluators, favoring policy evaluation, the transfer of information between different individuals and design the alternative that best suits the identified preferences. At the same time, with this information you can build communities of people with similar systems of preferences referred to a particular issue. The thesis shows a case of application of this methodology: optimization of labour policies in a real market. To be able support jobseekers (in their initiation or reinstatement to employment or when changing area of professional activity) is necessary to know their preferences for jobs that he is willing to perform. In addition, for labour mediation to be effective occupations that are sought must be offered by the labour market and the applicant must meet the conditions for access to these occupations. Further development of these models leads us to the procedures used to transform multiple preferences in an aggregate decision and consider both the views of each of the individuals involved in the decision and the social interactions, all aimed at generating a solution that best fits of the point of view of the entire population. Decisions with multiple participants mainly focus on: increasing the scope to include people who traditionally have not been considered in decision making, aggregation of the preferences of multiple participants in collective decision making (by vote, using applications developed for the Web 2.0 and through interpersonal comparisons of utility) and, finally, self-organization to allow participants to interact with each other in the assessment, so that the final result is better than the mere aggregation of individual opinions. The thesis analyzes the systems of e-democracy or tools for implementation which are more popular or more advanced. They are closely related to the Web 2.0 and its implementation is bringing an evolution of the current way of understanding democracy. We have also studied Collaborative Decision-Making (CDM)) software applications in decision-making which help to give sense and meaning to the data. They intend to coordinate the functions and features needed to reach adequate collective decisions, allowing all stakeholders to participate in the process. The thesis concludes with the presentation of a new model or paradigm in decision-making with multiple participants. The development is based on the calculation of the empathic utility functions. It seeks collaboration between individuals to make decision-making more effective; it also aims to increase the number of people involved. It studies the interactions and feedback among citizens, because the influence of some citizens in the other is fundamental to the process of collective decision-making and e-democracy. It also includes methods for detecting when the process has stalled and should be discontinued. This model is applied to the consultation of the citizens of a municipality on the opportunity to introduce bike lanes and characteristics they should have. Voting and interaction among voters is simulated.