Orientador: Ricardo da Silva Torres / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação / Made available in DSpace on 2018-08-13T11:00:01Z (GMT). No. of bitstreams: 1 Penatti_OtavioAugustoBizetto_M.pdf: 2250748 bytes, checksum: 57d5b2f9120a8eae69ee9881d363e9ce (MD5) Previous issue date: 2009 / Resumo: A crescente quantidade de imagens geradas e disponibilizadas atualmente tem eito aumentar a necessidade de criação de sistemas de busca para este tipo de informação. Um método promissor para a realização da busca de imagens e a busca por conteúdo. Este tipo de abordagem considera o conteúdo visual das imagens, como cor, textura e forma de objetos, para indexação e recuperação. A busca de imagens por conteúdo tem como componente principal o descritor de imagens. O descritor de imagens é responsável por extrair propriedades visuais das imagens e armazená-las em vetores de características. Dados dois vetores de características, o descritor compara-os e retorna um valor de distancia. Este valor quantifica a diferença entre as imagens representadas pelos vetores. Em um sistema de busca de imagens por conteúdo, a distancia calculada pelo descritor de imagens é usada para ordenar as imagens da base em relação a uma determinada imagem de consulta. Esta dissertação realiza um estudo comparativo de descritores de imagens considerando a Web como cenário de uso. Este cenário apresenta uma quantidade muito grande de imagens e de conteúdo bastante heterogêneo. O estudo comparativo realizado nesta dissertação é feito em duas abordagens. A primeira delas considera a complexidade assinto tica dos algoritmos de extração de vetores de características e das funções de distancia dos descritores, os tamanhos dos vetores de características gerados pelos descritores e o ambiente no qual cada descritor foi validado originalmente. A segunda abordagem compara os descritores em experimentos práticos em quatro bases de imagens diferentes. Os descritores são avaliados segundo tempo de extração, tempo para cálculos de distancia, requisitos de armazenamento e eficácia. São comparados descritores de cor, textura e forma. Os experimentos são realizados com cada tipo de descritor independentemente e, baseado nestes resultados, um conjunto de descritores é avaliado em uma base com mais de 230 mil imagens heterogêneas, que reflete o conteúdo encontrado na Web. A avaliação de eficácia dos descritores na base de imagens heterogêneas é realizada por meio de experimentos com usuários reais. Esta dissertação também apresenta uma ferramenta para a realização automatizada de testes comparativos entre descritores de imagens. / Abstract: The growth in size of image collections and the worldwide availability of these collections has increased the demand for image retrieval systems. A promising approach to address this demand is to retrieve images based on image content (Content-Based Image Retrieval). This approach considers the image visual properties, like color, texture and shape of objects, for indexing and retrieval. The main component of a content-based image retrieval system is the image descriptor. The image descriptor is responsible for encoding image properties into feature vectors. Given two feature vectors, the descriptor compares them and computes a distance value. This value quantifies the difference between the images represented by their vectors. In a content-based image retrieval system, these distance values are used to rank database images with respect to their distance to a given query image. This dissertation presents a comparative study of image descriptors considering the Web as the environment of use. This environment presents a huge amount of images with heterogeneous content. The comparative study was conducted by taking into account two approaches. The first approach considers the asymptotic complexity of feature vectors extraction algorithms and distance functions, the size of the feature vectors generated by the descriptors and the environment where each descriptor was validated. The second approach compares the descriptors in practical experiments using four different image databases. The evaluation considers the time required for features extraction, the time for computing distance values, the storage requirements and the effectiveness of each descriptor. Color, texture, and shape descriptors were compared. The experiments were performed with each kind of descriptor independently and, based on these results, a set of descriptors was evaluated in an image database containing more than 230 thousand heterogeneous images, reflecting the content existent in the Web. The evaluation of descriptors effectiveness in the heterogeneous database was made by experiments using real users. This dissertation also presents a tool for executing experiments aiming to evaluate image descriptors. / Mestrado / Sistemas de Informação / Mestre em Ciência da Computação

Science educators are currently facing the challenge of reforming the practices of science education. Publications of various science and educational organizations have established new criteria for accomplishing this goal. The new goal of science educators is scientific literacy for all.

User Modeling and Machine Learning for User Modeling have both become important research topics and key techniques in recent adaptive systems. One of the most intriguing problems in the `information age´ is how to filter relevant information from the huge amount of available data. This problem is tackled by using models of the user´s interest in order to increase precision and discriminate interesting information from un-interesting data. However, any user modeling approach suffers from several major drawbacks: User models built by the system need to be inspectable and understandable by the user himself. Secondly, users in general are not willing to give feedback concerning user satisfaction by the delivered results. Without any evidence for the user´s interest, it is hard to induce a hypothetical user model at all. Finally, most current systems do not draw a line of distinction between domain knowledge and user model which makes the adequacy of a user model hard to determine. This thesis presents the novel approach of conceptual user models. Conceptual user models are easy to inspect and understand and allow for the system to explain its actions to the user. It is shown, that ILP can be applied for the task of inducing user models from feedback, and a method for using mutual feedback for sample enlargement is introduced. Results are evaluated independently of domain knowledge within a clear machine learning problem definition. The whole concept presented is realized in a meta web search engine called OySTER.