Elements pour adapter les systemes de recherche d'information aux dyslexiques [Elements to adapt information retrieval systems to dyslexic users]

RÉSUMÉ. La prise en compte des troubles de la communication dans l’utilisation des systèmes de recherche d’information tels qu’on peut en trouver sur le Web est généralement réalisée par des interfaces utilisant des modalités n’impliquant pas la lecture et l’écriture. Peu d’applications existent pour aider l’utilisateur en difficulté dans la modalité textuelle. Nous proposons la prise en compte de la conscience phonologique pour assister l’utilisateur en difficulté d’écriture de requêtes (dysorthographie) ou de lecture de documents (dyslexie). En premier lieu un système de réécriture et d’interprétation des requêtes entrées au clavier par l’utilisateur est proposé : en s’appuyant sur les causes de la dysorthographie et sur les exemples à notre disposition, il est apparu qu’un système combinant une approche éditoriale (type correcteur orthographique) et une approche orale (système de transcription automatique) était plus approprié. En second lieu une méthode d’apprentissage automatique utilise des critères spécifiques , tels que la cohésion grapho-phonémique, pour estimer la lisibilité d’une phrase, puis d’un texte. ABSTRACT. Most applications intend to help disabled users in the information retrieval process by proposing non-textual modalities. This paper introduces specific parameters linked to phonological awareness in the textual modality. This will enhance the ability of systems to deal with orthographic issues and with the adaptation of results to the reader when for example the reader is dyslexic. We propose a phonology based sentence level rewriting system that combines spelling correction, speech synthesis and automatic speech recognition. This has been evaluated on a corpus of questions we get from dyslexic children. We propose a specific sentence readability measure that involves phonetic parameters such as grapho-phonemic cohesion. This has been learned on a corpus of reading time of sentences read by dyslexic children.

Fast and effective focused retrieval

Building an efficient and an effective search engine is a very challenging task. In this paper, we present the efficiency and effectiveness of our search engine at the INEX 2009 Efficiency and Ad Hoc Tracks. We have developed a simple and effective pruning method for fast query evaluation, and used a two-step process for Ad Hoc retrieval. The overall results from both tracks show that our search engine performs very competitively in terms of both efficiency and effectiveness.

Semantics-based web service discovery using information retrieval techniques

This paper demonstrates an experimental study that examines the accuracy of various information retrieval techniques for Web service discovery. The main goal of this research is to evaluate algorithms for semantic web service discovery. The evaluation is comprehensively benchmarked using more than 1,700 real-world WSDL documents from INEX 2010 Web Service Discovery Track dataset. For automatic search, we successfully use Latent Semantic Analysis and BM25 to perform Web service discovery. Moreover, we provide linking analysis which automatically links possible atomic Web services to meet the complex requirements of users. Our fusion engine recommends a final result to users. Our experiments show that linking analysis can improve the overall performance of Web service discovery. We also find that keyword-based search can quickly return results but it has limitation of understanding users’ goals.

Overview of the INEX 2010 focused relevance feedback track

The INEX 2010 Focused Relevance Feedback track offered a refined approach to the evaluation of Focused Relevance Feedback algorithms through simulated exhaustive user feedback. As in traditional approaches we simulated a user-in-the loop by re-using the assessments of ad-hoc retrieval obtained from real users who assess focused ad-hoc retrieval submissions. The evaluation was extended in several ways: the use of exhaustive relevance feedback over entire runs; the evaluation of focused retrieval where both the retrieval results and the feedback are focused; the evaluation was performed over a closed set of documents and complete focused assessments; the evaluation was performed over executable implementations of relevance feedback algorithms; and �finally, the entire evaluation platform is reusable. We present the evaluation methodology, its implementation, and experimental results obtained for nine submissions from three participating organisations.

A rank aggregation framework for video multimodal geocoding

This paper proposes a rank aggregation framework for video multimodal geocoding. Textual and visual descriptions associated with videos are used to define ranked lists. These ranked lists are later combined, and the resulting ranked list is used to define appropriate locations for videos. An architecture that implements the proposed framework is designed. In this architecture, there are specific modules for each modality (e.g, textual and visual) that can be developed and evolved independently. Another component is a data fusion module responsible for combining seamlessly the ranked lists defined for each modality. We have validated the proposed framework in the context of the MediaEval 2012 Placing Task, whose objective is to automatically assign geographical coordinates to videos. Obtained results show how our multimodal approach improves the geocoding results when compared to methods that rely on a single modality (either textual or visual descriptors). We also show that the proposed multimodal approach yields comparable results to the best submissions to the Placing Task in 2012 using no extra information besides the available development/training data. Another contribution of this work is related to the proposal of a new effectiveness evaluation measure. The proposed measure is based on distance scores that summarize how effective a designed/tested approach is, considering its overall result for a test dataset. © 2013 Springer Science+Business Media New York.

User-centric music information retrieval

The rapid growth of the Internet and the advancements of the Web technologies have made it possible for users to have access to large amounts of on-line music data, including music acoustic signals, lyrics, style/mood labels, and user-assigned tags. The progress has made music listening more fun, but has raised an issue of how to organize this data, and more generally, how computer programs can assist users in their music experience. An important subject in computer-aided music listening is music retrieval, i.e., the issue of efficiently helping users in locating the music they are looking for. Traditionally, songs were organized in a hierarchical structure such as genre->artist->album->track, to facilitate the users’ navigation. However, the intentions of the users are often hard to be captured in such a simply organized structure. The users may want to listen to music of a particular mood, style or topic; and/or any songs similar to some given music samples. This motivated us to work on user-centric music retrieval system to improve users’ satisfaction with the system. The traditional music information retrieval research was mainly concerned with classification, clustering, identification, and similarity search of acoustic data of music by way of feature extraction algorithms and machine learning techniques. More recently the music information retrieval research has focused on utilizing other types of data, such as lyrics, user-access patterns, and user-defined tags, and on targeting non-genre categories for classification, such as mood labels and styles. This dissertation focused on investigating and developing effective data mining techniques for (1) organizing and annotating music data with styles, moods and user-assigned tags; (2) performing effective analysis of music data with features from diverse information sources; and (3) recommending music songs to the users utilizing both content features and user access patterns.

Overview of the INEX 2011 relevance feedback track

The INEX 2011 Relevance Feedback track offered a refined approach to the evaluation of Focused Relevance Feedback algorithms through simulated exhaustive user feedback. Run in largely identical fashion to the Relevance Feedback track in INEX 2010[2], we simulated a user-in-the loop by re-using the assessments of ad-hoc retrieval obtained from real users who assess focused ad-hoc retrieval submissions. We present the evaluation methodology, its implementation, and experimental results obtained for four submissions from two participating organisations. As the task and evaluation methods did not change between INEX 2010 and now, explanations of these details from the INEX 2010 version of the track have been repeated verbatim where appropriate.

Overview of the INEX 2011 snippet retrieval track

This paper gives an overview of the INEX 2011 Snippet Retrieval Track. The goal of the Snippet Retrieval Track is to provide a common forum for the evaluation of the effectiveness of snippets, and to investigate how best to generate snippets for search results, which should provide the user with sufficient information to determine whether the underlying document is relevant. We discuss the setup of the track, and the evaluation results.

SRDigital: proposta de um modelo baseado na linguagem natural e controlada como instrumentos de apoio ao agente computacional do processo de referência

Pós-graduação em Ciência da Informação - FFC

Proposal of a trust and reputation model to provide recommendation in open quality oriented environments

El auge y penetración de las nuevas tecnologías junto con la llamada Web Social están cambiando la forma en la que accedemos a la medicina. Cada vez más pacientes y profesionales de la medicina están creando y consumiendo recursos digitales de contenido clínico a través de Internet, surgiendo el problema de cómo asegurar la fiabilidad de estos recursos. Además, un nuevo concepto está apareciendo, el de pervasive healthcare o sanidad ubicua, motivado por pacientes que demandan un acceso a los servicios sanitarios en todo momento y en todo lugar. Este nuevo escenario lleva aparejado un problema de confianza en los proveedores de servicios sanitarios. Las plataformas de eLearning se están erigiendo como paradigma de esta nueva Medicina 2.0 ya que proveen un servicio abierto a la vez que controlado/supervisado a recursos digitales, y facilitan las interacciones y consultas entre usuarios, suponiendo una buena aproximación para esta sanidad ubicua. En estos entornos los problemas de fiabilidad y confianza pueden ser solventados mediante la implementación de mecanismos de recomendación de recursos y personas de manera confiable. Tradicionalmente las plataformas de eLearning ya cuentan con mecanismos de recomendación, si bien están más enfocados a la recomendación de recursos. Para la recomendación de usuarios es necesario acudir a mecanismos más elaborados como son los sistemas de confianza y reputación (trust and reputation) En ambos casos, tanto la recomendación de recursos como el cálculo de la reputación de los usuarios se realiza teniendo en cuenta criterios principalmente subjetivos como son las opiniones de los usuarios. En esta tesis doctoral proponemos un nuevo modelo de confianza y reputación que combina evaluaciones automáticas de los recursos digitales en una plataforma de eLearning, con las opiniones vertidas por los usuarios como resultado de las interacciones con otros usuarios o después de consumir un recurso. El enfoque seguido presenta la novedad de la combinación de una parte objetiva con otra subjetiva, persiguiendo mitigar el efecto de posibles castigos subjetivos por parte de usuarios malintencionados, a la vez que enriquecer las evaluaciones objetivas con información adicional acerca de la capacidad pedagógica del recurso o de la persona. El resultado son recomendaciones siempre adaptadas a los requisitos de los usuarios, y de la máxima calidad tanto técnica como educativa. Esta nueva aproximación requiere una nueva herramienta para su validación in-silico, al no existir ninguna aplicación que permita la simulación de plataformas de eLearning con mecanismos de recomendación de recursos y personas, donde además los recursos sean evaluados objetivamente. Este trabajo de investigación propone pues una nueva herramienta, basada en el paradigma de programación orientada a agentes inteligentes para el modelado de comportamientos complejos de usuarios en plataformas de eLearning. Además, la herramienta permite también la simulación del funcionamiento de este tipo de entornos dedicados al intercambio de conocimiento. La evaluación del trabajo propuesto en este documento de tesis se ha realizado de manera iterativa a lo largo de diferentes escenarios en los que se ha situado al sistema frente a una amplia gama de comportamientos de usuarios. Se ha comparado el rendimiento del modelo de confianza y reputación propuesto frente a dos modos de recomendación tradicionales: a) utilizando sólo las opiniones subjetivas de los usuarios para el cálculo de la reputación y por extensión la recomendación; y b) teniendo en cuenta sólo la calidad objetiva del recurso sin hacer ningún cálculo de reputación. Los resultados obtenidos nos permiten afirmar que el modelo desarrollado mejora la recomendación ofrecida por las aproximaciones tradicionales, mostrando una mayor flexibilidad y capacidad de adaptación a diferentes situaciones. Además, el modelo propuesto es capaz de asegurar la recomendación de nuevos usuarios entrando al sistema frente a la nula recomendación para estos usuarios presentada por el modo de recomendación predominante en otras plataformas que basan la recomendación sólo en las opiniones de otros usuarios. Por último, el paradigma de agentes inteligentes ha probado su valía a la hora de modelar plataformas virtuales complejas orientadas al intercambio de conocimiento, especialmente a la hora de modelar y simular el comportamiento de los usuarios de estos entornos. La herramienta de simulación desarrollada ha permitido la evaluación del modelo de confianza y reputación propuesto en esta tesis en una amplia gama de situaciones diferentes. ABSTRACT Internet is changing everything, and this revolution is especially present in traditionally offline spaces such as medicine. In recent years health consumers and health service providers are actively creating and consuming Web contents stimulated by the emergence of the Social Web. Reliability stands out as the main concern when accessing the overwhelming amount of information available online. Along with this new way of accessing the medicine, new concepts like ubiquitous or pervasive healthcare are appearing. Trustworthiness assessment is gaining relevance: open health provisioning systems require mechanisms that help evaluating individuals’ reputation in pursuit of introducing safety to these open and dynamic environments. Technical Enhanced Learning (TEL) -commonly known as eLearning- platforms arise as a paradigm of this Medicine 2.0. They provide an open while controlled/supervised access to resources generated and shared by users, enhancing what it is being called informal learning. TEL systems also facilitate direct interactions amongst users for consultation, resulting in a good approach to ubiquitous healthcare. The aforementioned reliability and trustworthiness problems can be faced by the implementation of mechanisms for the trusted recommendation of both resources and healthcare services providers. Traditionally, eLearning platforms already integrate recommendation mechanisms, although this recommendations are basically focused on providing an ordered classifications of resources. For users’ recommendation, the implementation of trust and reputation systems appears as the best solution. Nevertheless, both approaches base the recommendation on the information from the subjective opinions of other users of the platform regarding the resources or the users. In this PhD work a novel approach is presented for the recommendation of both resources and users within open environments focused on knowledge exchange, as it is the case of TEL systems for ubiquitous healthcare. The proposed solution adds the objective evaluation of the resources to the traditional subjective personal opinions to estimate the reputation of the resources and of the users of the system. This combined measure, along with the reliability of that calculation, is used to provide trusted recommendations. The integration of opinions and evaluations, subjective and objective, allows the model to defend itself against misbehaviours. Furthermore, it also allows ‘colouring’ cold evaluation values by providing additional quality information such as the educational capacities of a digital resource in an eLearning system. As a result, the recommendations are always adapted to user requirements, and of the maximum technical and educational quality. To our knowledge, the combination of objective assessments and subjective opinions to provide recommendation has not been considered before in the literature. Therefore, for the evaluation of the trust and reputation model defined in this PhD thesis, a new simulation tool will be developed following the agent-oriented programming paradigm. The multi-agent approach allows an easy modelling of independent and proactive behaviours for the simulation of users of the system, conforming a faithful resemblance of real users of TEL platforms. For the evaluation of the proposed work, an iterative approach have been followed, testing the performance of the trust and reputation model while providing recommendation in a varied range of scenarios. A comparison with two traditional recommendation mechanisms was performed: a) using only users’ past opinions about a resource and/or other users; and b) not using any reputation assessment and providing the recommendation considering directly the objective quality of the resources. The results show that the developed model improves traditional approaches at providing recommendations in Technology Enhanced Learning (TEL) platforms, presenting a higher adaptability to different situations, whereas traditional approaches only have good results under favourable conditions. Furthermore the promotion period mechanism implemented successfully helps new users in the system to be recommended for direct interactions as well as the resources created by them. On the contrary OnlyOpinions fails completely and new users are never recommended, while traditional approaches only work partially. Finally, the agent-oriented programming (AOP) paradigm has proven its validity at modelling users’ behaviours in TEL platforms. Intelligent software agents’ characteristics matched the main requirements of the simulation tool. The proactivity, sociability and adaptability of the developed agents allowed reproducing real users’ actions and attitudes through the diverse situations defined in the evaluation framework. The result were independent users, accessing to different resources and communicating amongst them to fulfil their needs, basing these interactions on the recommendations provided by the reputation engine.

The mosaic test:benchmarking colour-based image retrieval systems using image mosaics

Evaluation and benchmarking in content-based image retrieval has always been a somewhat neglected research area, making it difficult to judge the efficacy of many presented approaches. In this paper we investigate the issue of benchmarking for colour-based image retrieval systems, which enable users to retrieve images from a database based on lowlevel colour content alone. We argue that current image retrieval evaluation methods are not suited to benchmarking colour-based image retrieval systems, due in main to not allowing users to reflect upon the suitability of retrieved images within the context of a creative project and their reliance on highly subjective ground-truths. As a solution to these issues, the research presented here introduces the Mosaic Test for evaluating colour-based image retrieval systems, in which test-users are asked to create an image mosaic of a predetermined target image, using the colour-based image retrieval system that is being evaluated. We report on our findings from a user study which suggests that the Mosaic Test overcomes the major drawbacks associated with existing image retrieval evaluation methods, by enabling users to reflect upon image selections and automatically measuring image relevance in a way that correlates with the perception of many human assessors. We therefore propose that the Mosaic Test be adopted as a standardised benchmark for evaluating and comparing colour-based image retrieval systems.

Link-the-Wiki : performance evaluation based on frequent phrases

In this paper, we discuss our participation to the INEX 2008 Link-the-Wiki track. We utilized a sliding window based algorithm to extract the frequent terms and phrases. Using the extracted phrases and term as descriptive vectors, the anchors and relevant links (both incoming and outgoing) are recognized efficiently.