iMente, servicios de información de actualidad en línea

iMente es un servicio de información de prensa digital realizado en España, que da acceso a los contenidos de publicaciones en línea que incluyen medios de comunicación, notas de prensa, weblogs y boletines oficiales. Se sitúa en el contexto de los productos de información periodística; se describen sus orígenes, evolución, tecnología, contenidos y tipos de usuarios; y se analizan sus principales prestaciones documentales, como seguimientos de prensa, alertas, búsquedas y publicación de titulares.

El negocio de buscar en internet. Análisis del mercado de los buscadores en 2003.

LOS BUSCADORES DE INTERNET se han convertido en pocos años en un punto de referencia imprescindible para la obtención de todo tipo de información. Representan las verdaderas páginas amarillas de la Red. Su grado de presencia en los hábitos telemáticos de la población es tan grande que resulta difícil imaginar una sesión cualquiera en internet sin la utilización de un buscador.

Búsqueda federada en el ecosistema de la e-ciencia: el caso Science Research

Se analiza el buscador Science Research en el contexto de la e-ciencia y de la búsqueda federada en comparación con la indización y la recolección. Se pone en relación también con otros buscadores académicos, especialmente Scirus y Google Scholar. Se realiza un análisis de los diversos componentes de Science Research y un estudio comparativo de obtención de resultados.

Valoracion del Impacto de la Informacion en Internet: Altavista, el 'Citation Index' de la red

En el presente artículo se recoge una metodología para la valoración del impacto de la información en Internet, usando las capacidades de indización y recuperación del buscador Altavista. Se aprovecha el contexto para describir la función de los metaelementos del HTML como mecanismo de estructuración y ordenación de la información. Se discuten las limitaciones y fiabilidad del método y se exponen algunos datos que muestran la producción de páginas WWW a nivel de institución y a nivel nacional, así como su comparación con otros países europeos. Se hace especial hincapié en la posibilidad de medir el impacto de estas páginas en función de las veces que son 'enlazadas' desde páginas externas de manera similar a como funciona el 'Citation Index' del Institute for Scientific Information.

Anàlisi del cercador múltiple Copernic 2001 Pro

Podeu consultar la versió en castellà a http://hdl.handle.net/2445/8965

Análisis del buscador múltiple Copernic 2001 Pro

Podeu consultar la versió en català a http://hdl.handle.net/2445/8964

Electronic Reference Library Phase I Report, 1999

This report describes the first phase in a project to develop an electronic reference library (ERL) to help Iowa transportation officials efficiently access information in critical and heavily used documents. These documents include Standard Specifications for Bridge and Highway Construction (hereinafter called Standard Specifications), design manuals, standard drawings, the Construction Manual, and Material Instruction Memoranda (hereinafter called Material IMs). Additional items that could be included to enhance the ERL include phone books, letting dates, Internet links, computer programs distributed by the Iowa Department of Transportation (DOT), and local specifications, such as the Urban Standard Specifications of Public Improvements. All cross-references should be hyper linked, and a search engine should be provided. Revisions noted in the General Supplemental Specifications (hereinafter called the Supplemental Specifications) should be incorporated into the text of the Standard Specifications. The Standard Specifications should refer to related sections of other documents, and there should be reciprocal hyper links in those other documents. These features would speed research on critical issues and save staff time. A master plan and a pilot version were both developed in this first phase of the ERL.

FOSS License Selection and Code Management

With nearly 2,000 free and open source software (FLOSS) licenses, software license proliferation¿ can be a major headache for software development organizations trying to speed development through software component reuse, as well as companies redistributing software packages as components of their products. Scope is one problem: from the Free Beer license to the GPL family of licenses to platform-specific licenses such as Apache and Eclipse, the number and variety of licenses make it difficult for companies to ¿do the right thing¿ with respect to the software components in their products and applications. In addition to the sheer number of licenses, each license carries within it the author¿s specific definition of how the software can be used and re-used. Permissive licenses like BSD and MIT make it easy; software can be redistributed and developers can modify code without the requirement of making changes publicly available. Reciprocal licenses, on the other hand, place varying restrictions on re-use and redistribution. Woe to the developer who snags a bit of code after a simple web search without understanding the ramifications of license restrictions.

La cerca única, o com fer fàcil el camí a l'usuari

This paper presents a reflection on the need for libraries to think about how to facilitate access to the documentary sources they manage.As the number of resources available in electronic form increases, libraries are in the need to provide a simple and usable search tool that allows integrating the contents of the various information management systems they give access to.To define user expectations to the search interface, some of the features that they are accustomed to use in their requests for information on the Internet have been included.The technologies that allow the discovery layer implementation as a search tool that integrates the various information systems of the library are presented next. And below are some examples of implementations that work in line with the integration of various information sources into a single search engine, as models to consider for implementing a system of this kind.The purpose of it all is to present a state of the art of some cases of operational deployments as a starting point for any organization interested in improving access it offers to its resources on the basis of references study.

The Achieved Dreams of a Neighborhood

This article explains the social transformation process initiated at the end of the 1970s within the neighborhood of La Verneda-Sant Martı´ in Barcelona. This process started with the foundation of an adult education center that was organized as a Learning Community (the first one in the world). From the beginning, it was administered for and by the community. It became a space of debate where the demands and dreams of the neighbors converged about transforming their neighborhood along with the recommendations of the international scientific community. Twenty years later, the dreams came true: There have been substantial improvements throughout the urban space, infrastructures, housing, urban thoroughfares, and public highways. The INCLUD-ED European project, using the communicative methodology of research, has thoroughly studied the transformation carried out in the La Verneda-Sant Martı´ Adult School and its neighborhood. INCLUD-ED has identified successful practices within diverse social areas that are transferable to other contexts and contribute to overcoming inequalities and improving the most underprivileged neighborhoods.

Àgora. La experiencia de la Escuela de personas adultas. La verneda-Sant Martí: del aprendizaje de las tecnologías al contexto de aprendizaje tecnológico

La Comunidad de Aprendizaje Escuela de personas adultas La Verneda-Sant Martí, lleva más de 10 años trabajando las tecnologías de la información y la comunicación desde una perspectiva transversal y global. A través del trabajo que venimos realizando día a día, hemos visto como las TIC han pasado de ser una necesidad formativa a convertirse en un contexto de aprendizaje cotidiano entre las personas que participan de nuestro proyecto. Desde que en 1999 decidimos transformar nuestras actividades para integrarnos plenamente en la sociedad de la información hemos cambiado metodologías, prioridades y formas de trabajar. Mostrar cuáles han sido esas transformaciones y qué resultados son los que se han alcanzado es el principal reto del artículo.

Unsupervised visual object categorization

The number of digital images has been increasing exponentially in the last few years. People have problems managing their image collections and finding a specific image. An automatic image categorization system could help them to manage images and find specific images. In this thesis, an unsupervised visual object categorization system was implemented to categorize a set of unknown images. The system is unsupervised, and hence, it does not need known images to train the system which needs to be manually obtained. Therefore, the number of possible categories and images can be huge. The system implemented in the thesis extracts local features from the images. These local features are used to build a codebook. The local features and the codebook are then used to generate a feature vector for an image. Images are categorized based on the feature vectors. The system is able to categorize any given set of images based on the visual appearance of the images. Images that have similar image regions are grouped together in the same category. Thus, for example, images which contain cars are assigned to the same cluster. The unsupervised visual object categorization system can be used in many situations, e.g., in an Internet search engine. The system can categorize images for a user, and the user can then easily find a specific type of image.

In Google we trust?

In a micro-founded model, we derive novel incentives for a monopoly search engine to distort its organic and its sponsored results on searches for online content and offline products. Distorting organic results towards content publishers with less effective display advertising and/or distorting sponsored results towards higher margin merchants (by underweighting consumer relevance in search auctions) increase per capita revenues but lower participation. The interplay of these incentives determines search bias and welfare. We also characterize how the welfare consequences of integration into display advertising, as intermediary or publisher, depend on asymmetries, monopolization and targeting.

Kernel-Based Ranking. Methods for Learning and Performance Estimation

Machine learning provides tools for automated construction of predictive models in data intensive areas of engineering and science. The family of regularized kernel methods have in the recent years become one of the mainstream approaches to machine learning, due to a number of advantages the methods share. The approach provides theoretically well-founded solutions to the problems of under- and overfitting, allows learning from structured data, and has been empirically demonstrated to yield high predictive performance on a wide range of application domains. Historically, the problems of classification and regression have gained the majority of attention in the field. In this thesis we focus on another type of learning problem, that of learning to rank. In learning to rank, the aim is from a set of past observations to learn a ranking function that can order new objects according to how well they match some underlying criterion of goodness. As an important special case of the setting, we can recover the bipartite ranking problem, corresponding to maximizing the area under the ROC curve (AUC) in binary classification. Ranking applications appear in a large variety of settings, examples encountered in this thesis include document retrieval in web search, recommender systems, information extraction and automated parsing of natural language. We consider the pairwise approach to learning to rank, where ranking models are learned by minimizing the expected probability of ranking any two randomly drawn test examples incorrectly. The development of computationally efficient kernel methods, based on this approach, has in the past proven to be challenging. Moreover, it is not clear what techniques for estimating the predictive performance of learned models are the most reliable in the ranking setting, and how the techniques can be implemented efficiently. The contributions of this thesis are as follows. First, we develop RankRLS, a computationally efficient kernel method for learning to rank, that is based on minimizing a regularized pairwise least-squares loss. In addition to training methods, we introduce a variety of algorithms for tasks such as model selection, multi-output learning, and cross-validation, based on computational shortcuts from matrix algebra. Second, we improve the fastest known training method for the linear version of the RankSVM algorithm, which is one of the most well established methods for learning to rank. Third, we study the combination of the empirical kernel map and reduced set approximation, which allows the large-scale training of kernel machines using linear solvers, and propose computationally efficient solutions to cross-validation when using the approach. Next, we explore the problem of reliable cross-validation when using AUC as a performance criterion, through an extensive simulation study. We demonstrate that the proposed leave-pair-out cross-validation approach leads to more reliable performance estimation than commonly used alternative approaches. Finally, we present a case study on applying machine learning to information extraction from biomedical literature, which combines several of the approaches considered in the thesis. The thesis is divided into two parts. Part I provides the background for the research work and summarizes the most central results, Part II consists of the five original research articles that are the main contribution of this thesis.

Inbound marketing in practice. Case study: Lainaaja.fi

The general purpose of the thesis was to describe and explain the particularities of inbound marketing methods and the key advantages of those methods. Inbound marketing can be narrowed down to a set of marketing strategies and techniques focused on pulling prospects towards a business and its products on the Internet by producing useful and relevant content to prospects. The main inbound marketing methods and channels were identified as blogging, content publishing, search engine optimization and social media. The best way to utilise these methods is producing great content that should cover subjects that interest the target group, which is usually a composition of buyers, existing customers and influencers, such as analysts and media The study revealed increase in Lainaaja.fi traffic and referral traffic sources that was firmly confirmed as statistically significant, while number of backlinks and SERP placement were clearly positively correlated, but not statistically significant. The number of new registered users along with new loan applicants and deposits did not show correlation with increased content producing. The conclusion of the study shows inbound marketing campaign clearly increasing website traffic and plausible help on getting better search engine results compared to control period. Implications are clear; inbound marketing is an activity that every business should consider implementing. But just producing content online is not enough; equal amount of work should be put into turning the visitors into customers. Further studies are recommended on using inbound marketing combined with monitoring of landing pages and conversion optimization to incoming visitors.