The use of visual data representation is increasing the possibilities to exchange information and communicate indifferent contexts all over the world. Communicating food wastage visually to influence consuming patterns isone of these possibilities. Food wastage is currently a much-prioritized topic in Sweden as well as globally due toits negative impacts on society, environment and the economy, and therefore there is much need to bringinnovative solutions supporting reduction of food waste. This thesis presents a qualitative research based on a casestudy of food waste management at Sala municipality in Sweden while exploring the current visual datarepresentation techniques and its further potential to make food waste management more sustainable. The researchframework used in this thesis is based on visual rhetoric and the innovation theories. The thesis analyzes foodwastage from an international perspective, its connection to sustainable development goals and how MatomaticAB uses a visual data representation tool to address food wastage.The thesis further explains how the users associated with Sala municipality interpret the existing tool, thechallenges they face and review their expectations to build a new visual data representation model. The results ofquestionnaires filled by user’s, state that 50% of the respondents understand the current tool to its full capacityand only 50% of the respondents are satisfied with the overall tool. When it comes to the choice of datapresentation 67% of the users showed interest in use of infographics instead of the conventional bar graphs, andtherefore some parameters like, making the tool more interesting using infographics, user friendly by limiting thedata displayed and interactive by giving user options to explore further as per their liking, were thought whiledesigning the new visual data representation model.

Background and Objective In stereotactic brain tumor biopsies, the combination of real-time fluorescence spectroscopy with the detection of microvascular perfusion using laser Doppler flowmetry provides an improved localization of the brain tumor while decreasing the risk of intra-cranial hemorrhage. The surgeon using the measurement probe is required to view signal values on a screen or usually, when her or his visual focus is directed at the patient, the verbal feedback of a biomedical engineer who is monitoring the measurement signals is needed. In this process possible important information can be overlooked and time is lost. The aim of the thesis was the development a visual and auditory user interface (UI) for use in stereotactic brain tumor biopsies. Materials and Methods The system translates the fluorescence intensity of protoporphyrin IX (PpIX) into sound and visual indicators that are easy and fast to recognize and transmits warning signals in case of signal error or the detection of microvascular perfusion. The increasing and de-creasing fluorescence values at tumor margins were reproduced to improve the precision of de-tecting varying fluorescence intensities when entering tumor tissue with color gradient models. The algorithm produced five signal values when specific fluorescence intensities were measured and compared at different wavelengths.For the development of the UI, a user-centered design was implemented. The user-, operating room- and safety requirements were gathered by communicating with the biomedical engineers and neurosurgeons who had experience in working with fluorescence guided brain tumor biop-sies. The requirements were considered when designing the UI’s features in LabVIEW and the auditory feedback was generated using OSC (Open Sound Control). The user interface intended to deliver measurement data to the user that triggered a high response accuracy by being easy to understand while inducing high user acceptance. The user interaction and function response accuracy of the visual and auditory interface were evaluated in statistical tests where operating room situations were mimicked. The user acceptance of the UI was evaluated. Results Signals for no, low (increasing and decreasing) and high fluorescence indicators, as well as two warning indicators for a blocked signal and vessel occurrence were represented visually and auditorially by the user interface. An intensity/time graph and intensity/wavelength graph, along with the option of recording measurement files and opening saved files allowed the inspec-tion of detailed measurement values. The user study exhibited auditory response accuracy of 95 ± 3% in the intuition test and 91±16% in a memory test. The testing of the response accuracy of the individual signal values displayed accurate responses in 84% to 100% of times a signal was played back. The user acceptance rating of the auditory and visual interface showed no negative results. Conclusion A UI was developed to visually and auditorially represent measurement values to a neurosurgeon performing a stereotactic brain tumor biopsy procedure and biomedical engineers monitoring the measurement signals. The visual display was successful in representing data in a way that was easy to understand. The auditory interface showed high response accuracies for the individual tones representing measurement values. The majority of the test subjects per-ceived the signals to be intuitive, easy to understand and easy to remember. The auditory and visual UI showed high user acceptance ratings, indicating that the user interface was useful and satisfactory in its application.

Aprendizado múltipla instância (MIL) é um paradigma de aprendizado de máquina que tem o objetivo de classificar um conjunto (bags) de objetos (instâncias), atribuindo rótulos só para os bags. Em MIL apenas os rótulos dos bags estão disponíveis para treinamento, enquanto os rótulos das instâncias são desconhecidos. Este problema é frequentemente abordado através da seleção de uma instância para representar cada bag, transformando um problema MIL em um problema de aprendizado supervisionado padrão. No entanto, não se conhecem abordagens que apoiem o usuário na realização desse processo. Neste trabalho, propomos uma visualização baseada em árvore multi-escala chamada MILTree que ajuda os usuários na realização de tarefas relacionadas com MIL, e também dois novos métodos de seleção de instâncias, chamados MILTree-SI e MILTree-Med, para melhorar os modelos MIL. MILTree é um layout de árvore de dois níveis, sendo que o primeiro projeta os bags, e o segundo nível projeta as instâncias pertencentes a cada bag, permitindo que o usuário explore e analise os dados multi-instância de uma forma intuitiva. Já os métodos de seleção de instãncias objetivam definir uma instância protótipo para cada bag, etapa crucial para a obtenção de uma alta precisão na classificação de dados multi-instância. Ambos os métodos utilizam o layout MILTree para atualizar visualmente as instâncias protótipo, e são capazes de lidar com conjuntos de dados binários e multi-classe. Para realizar a classificação dos bags, usamos um classificador SVM (Support Vector Machine). Além disso, com o apoio do layout MILTree também pode-se atualizar os modelos de classificação, alterando o conjunto de treinamento, a fim de obter uma melhor classificação. Os resultados experimentais validam a eficácia da nossa abordagem, mostrando que a mineração visual através da MILTree pode ajudar os usuários em cenários de classificação multi-instância. / Multiple-instance learning (MIL) is a paradigm of machine learning that aims at classifying a set (bags) of objects (instances), assigning labels only to the bags. In MIL, only the labels of bags are available for training while the labels of instances in bags are unknown. This problem is often addressed by selecting an instance to represent each bag, transforming a MIL problem into a standard supervised learning. However, there is no user support to assess this process. In this work, we propose a multi-scale tree-based visualization called MILTree that supports users in tasks related to MIL, and also two new instance selection methods called MILTree-SI and MILTree-Med to improve MIL models. MILTree is a two-level tree layout, where the first level projects bags, and the second level projects the instances belonging to each bag, allowing the user to understand the data multi-instance in an intuitive way. The developed selection methods define instance prototypes of each bag, which is important to achieve high accuracy in multi-instance classification. Both methods use the MILTree layout to visually update instance prototypes and can handle binary and multiple-class datasets. In order to classify the bags we use a SVM classifier. Moreover, with support of MILTree layout one can also update the classification model by changing the training set in order to obtain a better classifier. Experimental results validate the effectiveness of our approach, showing that visual mining by MILTree can help the users in MIL classification scenarios.

As técnicas de projeção ou posicionamento de pontos no plano, que servem para mapear dados multi-dimensionais em espaços visuais, sempre despertaram grande interesse da comunidade de visualização e análise de dados por representarem uma forma útil de exploração baseada em relações de similaridade e correlação. Apesar disso, muitos problemas ainda são encontrados em tais técnicas, limitando suas aplicações. Em especial, as técnicas de projeção multi-dimensional de maior qualidade têm custo computacional proibitivo para grandes conjuntos de dados. Adicionalmente, problemas referentes à escalabilidade visual, isto é, à capacidade da metáfora visual empregada de representar dados de forma compacta e amigável, são recorrentes. Esta tese trata o problema da projeção multi-dimensional de vários pontos de vista, propondo técnicas que resolvem, até certo ponto, cada um dos problemas verificados. Também é fato que a complexidade e o tamanho dos conjuntos de dados indicam que a visualização deve trabalhar em conjunto com técnicas de mineração, tanto embutidas no processo de mapeamento, como por meio de ferramentas auxiliares de interpretação. Nesta tese incorporamos alguns aspectos de mineração integrados ao processo de visualização multi-dimensional, principalmente na aplicação de projeções para visualização de coleções de documentos, propondo uma estratégia de extração de tópicos. Como suporte ao desenvolvimento e teste dessas técnicas, foram criados diferentes sistemas de software. O principal inclui as técnicas desenvolvidas e muitas das técnicas clássicas de projeção, podendo ser usado para exploração de conjuntos de dados multi-dimensionais em geral, com funcionalidade adicional para mapeamento de coleções de documentos. Como principal contribuição desta tese propomos um entendimento mais profundo dos problemas encontrados nas técnicas de projeção vigentes e o desenvolvimento de técnicas de projeção (ou mapeamento) que são rápidas, tratam adequadamente a formação visual de grupos de dados altamente similares, separam satisfatoriamente esses grupos no layout, e permitem a exploração dos dados em vários níveis de detalhe / Projection or point placement techniques, useful for mapping multidimensional data into visual spaces, have always risen interest in the visualization and data analysis communities because they can support data exploration based on similarity or correlation relations. Regardless of that interest, various problems arise when dealing with such techniques, impairing their widespread application. In particularly the projections that yield highest quality layouts have prohibitive computational cost for large data sets. Additionally, there are issues regarding visual scalability, i.e., the capability of visually fit the individual points in the exploration space as the data set grows large. This thesis treats the problems of projections from various perspectives, presenting novel techniques that solve, to certain extent, several of the verified problems. It is also a fact that size and complexity of data sets suggest the integration of data mining capabilities into the visualization pipeline, both during the mapping process and as a tools to extract additional information after the data have been layed out. This thesis also add some aspects of mining to the multidimensional visualization process, mainly for the particular application of analysis of document collections, proposing and implementing an approach for topic extraction. As supporting tools for testing these techniques and comparing them to existing ones different software systems were written. The main one includes the techniques developed here as well as several of the classical projection and dimensional reduction techniques, and can be used for exploring various kinds of data sets, with addition functionality to support the mapping of document collections. This thesis contributes to the understanding of the projection or mapping problem and develops new techniques that are fast, treat adequately the visual formation of groups of highly related data items, separate those groups properly and allow exploration of data in various levels of detail

Usuários da Internet estão muito familiarizados que resultados de uma consulta sejam exibidos como uma lista ordenada de snippets. Cada snippet possui conteúdo textual que mostra um resumo do documento referido (ou página web) e um link para o mesmo. Esta representação tem muitas vantagens como, por exemplo, proporcionar uma navegação fácil e simples de interpretar. No entanto, qualquer usuário que usa motores de busca poderia reportar possivelmente alguma experiência de decepção com este modelo. Todavia, ela tem limitações em situações particulares, como o não fornecimento de uma visão geral da coleção de documentos recuperados. Além disso, dependendo da natureza da consulta - por exemplo, pode ser muito geral, ou ambígua, ou mal expressa - a informação desejada pode ser mal classificada, ou os resultados podem contemplar temas variados. Várias tarefas de busca seriam mais fáceis se fosse devolvida aos usuários uma visão geral dos documentos organizados de modo a refletir a forma como são relacionados, em relação ao conteúdo. Propomos uma técnica de visualização para exibir os resultados de consultas web que visa superar tais limitações. Ela combina a capacidade de preservação de vizinhança das projeções multidimensionais com a conhecida representação baseada em snippets. Essa visualização emprega uma projeção multidimensional para derivar layouts bidimensionais dos resultados da pesquisa, que preservam as relações de similaridade de texto, ou vizinhança. A similaridade é calculada mediante a aplicação da similaridade do cosseno sobre uma representação bag-of-words vetorial de coleções construídas a partir dos snippets. Se os snippets são exibidos diretamente de acordo com o layout derivado, eles se sobrepõem consideravelmente, produzindo uma visualização pobre. Nós superamos esse problema definindo uma energia funcional que considera tanto a sobreposição entre os snippets e a preservação da estrutura de vizinhanças como foi dada no layout da projeção. Minimizando esta energia funcional é fornecida uma representação bidimensional com preservação das vizinhanças dos snippets textuais com sobreposição mínima. A visualização transmite tanto uma visão global dos resultados da consulta como os agrupamentos visuais que refletem documentos relacionados, como é ilustrado em vários dos exemplos apresentados / Internet users are very familiar with the results of a search query displayed as a ranked list of snippets. Each textual snippet shows a content summary of the referred document (or web page) and a link to it. This display has many advantages, e.g., it affords easy navigation and is straightforward to interpret. Nonetheless, any user of search engines could possibly report some experience of disappointment with this metaphor. Indeed, it has limitations in particular situations, as it fails to provide an overview of the document collection retrieved. Moreover, depending on the nature of the query - e.g., it may be too general, or ambiguous, or ill expressed - the desired information may be poorly ranked, or results may contemplate varied topics. Several search tasks would be easier if users were shown an overview of the returned documents, organized so as to reflect how related they are, content-wise. We propose a visualization technique to display the results of web queries aimed at overcoming such limitations. It combines the neighborhood preservation capability of multidimensional projections with the familiar snippet-based representation by employing a multidimensional projection to derive two-dimensional layouts of the query search results that preserve text similarity relations, or neighborhoods. Similarity is computed by applying the cosine similarity over a bag-of-words vector representation of collection built from the snippets. If the snippets are displayed directly according to the derived layout they will overlap considerably, producing a poor visualization. We overcome this problem by defining an energy functional that considers both the overlapping amongst snippets and the preservation of the neighborhood structure as given in vii the projected layout. Minimizing this energy functional provides a neighborhood preserving two-dimensional arrangement of the textual snippets with minimum overlap. The resulting visualization conveys both a global view of the query results and visual groupings that reflect related results, as illustrated in several examples shown

Nas últimas décadas, a capacidade das empresas de gerar e coletar informações aumentou rapidamente. Essa explosão no volume de dados gerou a necessidade do desenvolvimento de novas técnicas e ferramentas que pudessem, além de processar essa enorme quantidade de dados, permitir sua análise para a descoberta de informações úteis, de maneira inteligente e automática. Isso fez surgir um proeminente campo de pesquisa para a extração de informação em bases de dados denominado Knowledge Discovery in Databases – KDD, no geral técnicas de mineração de dados – DM – têm um papel preponderante. A obtenção de bons resultados na etapa de mineração de dados depende fortemente de quão adequadamente o preparo dos dados é realizado. Sendo assim, a etapa de extração de conhecimento (DM) no processo de KDD, é normalmente precedida de uma etapa de pré-processamento, onde os dados que porventura devam ser submetidos à etapa de DM são integrados em uma única relação. Um problema importante enfrentado nessa etapa é que, na maioria das vezes, o usuário ainda não tem uma idéia muito precisa dos dados que devem ser extraídos. Levando em consideração a grande habilidade de exploração da mente humana, este trabalho propõe uma técnica de visualização de dados armazenados em múltiplas relações de uma base de dados relacional, com o intuito de auxiliar o usuário na preparação dos dados a serem minerados. Esta técnica permite que a etapa de DM seja aplicada sobre múltiplas relações simultaneamente, trazendo as operações de junção para serem parte desta etapa. De uma maneira geral, a adoção de junções em ferramentas de DM não é prática, devido ao alto custo computacional associado às operações de junção. Entretanto, os resultados obtidos nas avaliações de desempenho da técnica proposta neste trabalho mostraram que ela reduz esse custo significativamente, tornando possível a exploração visual de múltiplas relações de uma maneira interativa. / In the last decades the capacity of information generation and accumulation increased quickly. With the explosive growth in the volume of data, new techniques and tools are being sought to process it and to automatically discover useful information from it, leading to techniques known as Knowledge Discovery in Databases – KDD – where, in general, data mining – DM – techniques play an important role. The results of applying data mining techniques on datasets are highly dependent on proper data preparation. Therefore, in traditional DM processes, data goes through a pre-processing step that results in just one table that is submitted to mining. An important problem faced during this step is that, most of the times, the analyst doesn’t have a clear idea of what portions of data should be mined. This work reckons the strong ability of human beings to interpret data represented in graphical format, to develop a technique to visualize data from multiple tables, helping human analysts when preparing data to DM. This technique allows the data mining process to be applied over multiple relations at once, bringing the join operations to become part of this process. In general, the use of multiple tables in DM tools is not practical, due to the high computational cost required to explore them. Experimental evaluation of the proposed technique shows that it reduces this cost significantly, turning it possible to visually explore data from multiple tables in an interactive way.

Esta tese apresenta um arcabouço teórico para auxiliar o estudo e o projeto de técnicas de visualização interativa de dados. Tais técnicas, tradicionalmente, têm sido projetadas baseando-se na experiência dos analistas desenvolvedores. Muitos trabalhos, todavia, têm procurado desenvolver um espaço de compreensão coerente para explicar como as visualizações são compostas e para permitir a predição de novas abordagens para técnicas de visualização. No entanto, propostas precursoras apresentam inadequações, não sendo capazes nem de fomentar novas sistematizações nem de explicar a concepção das técnicas mais recentes encontradas na literatura. Numa etapa inicial, esta tese revê conceitos em visualização, percepção e cognição procurando explicar como a análise visual de dados funciona. A revisão destes trabalhos é sintetizada em um processo de expressividade visual que correlaciona estímulos pré-atentivos, percepção visual analítica e interpretação cognitiva. Em seguida, após uma extensa revisão de trabalhos relacionados, a discussão prossegue definindo um plano de teorização da constituição dos métodos de representação visual de dados. Este plano impulsiona o desenvolvimento de uma sistematização inicial na forma de uma taxonomia capaz de caracterizar os constituintes pré-atentivos das visualizações. Esta caracterização é orientada à percepção visual analítica, que é parte do processo de expressividade visual. Desta maneira, em uma abordagem orientada a percepções visuais, as técnicas de visualização são classificadas de acordo com um conjunto limitado de características comuns e de processos de espacialização de dados. O próximo passo da discussão prossegue para a construção de um espaço de design com dimensões de posição, forma e cor. O espaço proposto, denominado Espaço de Design Espacial- Perceptivo, considera a possibilidade de múltiplos ciclos de espacialização de dados e também técnicas de interação. Baseando-se no espaço de design introduzido, a tese apresenta um modelo para a definição de parâmetros para o design de visualizações. Este modelo, que é um primeiro resultado da aplicação dos conceitos apresentados, prevê uma ferramenta para a definição, apresentação automática e avaliação empírica de representações visuais de dados. O trabalho é encerrado com a descrição de dois sistemas completos para a visualização de grafos e de dados multi variados. Assim, na última parte do texto, os sistemas GMine e VisTree são formalmente apresentados e analisados como estudos de caso à luz da teoria desenvolvida na tese / This thesis presents a theoretical framework to assist the study and the design of interactive data visualization techniques. Traditionally, visualization techniques have been designed based on analysts? experience. Many works, though, have sought to develop a coherent comprehension space to explain how visualizations are composed and to allow the prediction of new approaches for visualization techniques. However, precursor proposals present inadequacies and have not been able neither to furnish new systematizations nor to explain late techniques found in literature. In an initial step, this thesis reviews concepts on visualization, perception and cognition aiming at explaining how visual data analyses work. The revision of these works is synthesized in a process of visual expressivity that interrelates pre-attentive stimuli, analytical visual perception and cognitive interpretation. Then, after an extensive revision of related works, the discussion proceeds by structuring a plan for theorizing the constitution of methods for data visual representation. This plan furnishes the development of an initial systematization in the form of a taxonomy that characterizes the pre-attentive constituents of visualizations. This characterization considers visual analytical perceptions, which are part of the processes of visual expressivity. Like so, in a perceptions oriented approach, visualization techniques are classified according to a limited set of common characteristics and to data spatialization processes. The next step in the discussion proceeds to the construction of a space with dimensions position, shape and color. The proposed space is named Spatial/Perceptual Design Space, it considers the possibility of multiple cycles of data spatialization and also interaction techniques. Based on the design space just introduced, this thesis presents a model for the definition of parameters for visualization design. This model, which is a first result of the application of the presented concepts, foresees a tool for the definition, automatic presentation and empirical evaluation of visual data representations. The work is finished with the description of two complete systems for the visualization of graphs and multivariate data. Hence, in the last part of the text, systems GMine and VisTree are formally presented and analyzed as study cases under the light of the theory introduced in the thesis

Most of the largest databases currently available have a strong geospatial component and contain potentially useful information which might be of value. The discipline concerned with extracting this information and knowledge is data mining. Knowledge discovery is performed by applying automatic algorithms which recognise patterns in the data. Classical data mining algorithms assume that data are independently generated and identically distributed. Geospatial data are multidimensional, spatially autocorrelated and heterogeneous. These properties make classical data mining algorithms inappropriate for geospatial data, as their basic assumptions cease to be valid. Extracting knowledge from geospatial data therefore requires special approaches. One way to do that is to use visual data mining, where the data is presented in visual form for a human to perform the pattern recognition. When visual mining is applied to geospatial data, it is part of the discipline called exploratory geovisualisation. Both automatic and visual data mining have their respective advantages. Computers can treat large amounts of data much faster than humans, while humans are able to recognise objects and visually explore data much more effectively than computers. A combination of visual and automatic data mining draws together human cognitive skills and computer efficiency and permits faster and more efficient knowledge discovery. This thesis investigates if a combination of visual and automatic data mining is useful for exploration of geospatial data. Three case studies illustrate three different combinations of methods. Hierarchical clustering is combined with visual data mining for exploration of geographical metadata in the first case study. The second case study presents an attempt to explore an environmental dataset by a combination of visual mining and a Self-Organising Map. Spatial pre-processing and visual data mining methods were used in the third case study for emergency response data. Contemporary system design methods involve user participation at all stages. These methods originated in the field of Human-Computer Interaction, but have been adapted for the geovisualisation issues related to spatial problem solving. Attention to user-centred design was present in all three case studies, but the principles were fully followed only for the third case study, where a usability assessment was performed using a combination of a formal evaluation and exploratory usability. / QC 20110118

The large and ever-increasing amounts of multi-dimensional, multivariate, multi-source, spatio-temporal data represent a major challenge for the future. The need to analyse and make decisions based on these data streams, often in time-critical situations, demands integrated, automatic and sophisticated interactive tools that aid the user to manage, process, visualize and interact with large data spaces. The rise of `Web 2.0', which is undisputedly linked with developments such as blogs, wikis and social networking, and the internet usage explosion in the last decade represent another challenge for adapting these tools to the Internet to reach a broader user community. In this context, the research presented in this thesis introduces an effective web-enabled geovisual analytics framework implemented, applied and verified in Adobe Flash ActionScript and HTML5/JavaScript. It has been developed based on the principles behind Visual Analytics and designed to significantly reduce the time and effort needed to develop customized web-enabled applications for geovisual analytics tasks and to bring the benefits of visual analytics to the public. The framework has been developed based on a component architecture and includes a wide range of visualization techniques enhanced with various interaction techniques and interactive features to support better data exploration and analysis. The importance of multiple coordinated and linked views is emphasized and a number of effective techniques for linking views are introduced. Research has so far focused more on tools that explore and present data while tools that support capturing and sharing gained insight have not received the same attention. Therefore, this is one of the focuses of the research presented in this thesis. A snapshot technique is introduced, which supports capturing discoveries made during the exploratory data analysis process and can be used for sharing gained knowledge. The thesis also presents a number of applications developed to verify the usability and the overall performance of the framework for the visualization, exploration and analysis of data in different domains. Four application scenarios are presented introducing (1) the synergies among information visualization methods, geovisualization methods and volume data visualization methods for the exploration and correlation of spatio-temporal ocean data, (2) effective techniques for the visualization, exploration and analysis of self-organizing network data, (3) effective flow visualization techniques applied to the analysis of time-varying spatial interaction data such as migration data, commuting data and trade flow data, and (4) effective techniques for the visualization, exploration and analysis of flood data.

In this project, a prototype visual data mining tool was developed, allowing users to interactively investigate large multi-dimensional datasets visually (using 2D visualization techniques) using so called drill-down, roll-up and slicing operations. The project included all steps of the development, from writing specifications and designing the program to implementing and evaluating it. Using ideas from data warehousing, custom methods for storing pre-computed aggregations of data (commonly referred to as materialized views) and retrieving data from these were developed and implemented in order to achieve higher performance on large datasets. View materialization enables the program to easily fetch or calculate a view using other views, something which can yield significant performance gains if view sizes are much smaller than the underlying raw dataset. The choice of which views to materialize was done in an automated manner using a well-known algorithm - the greedy algorithm for view materialization - which selects the fraction of all possible views that is likely (but not guaranteed) to yield the best performance gain. The use of materialized views was shown to have good potential to increase performance for large datasets, with an average speedup (compared to on-the-fly queries) between 20 and 70 for a test dataset containing 500~000 rows. The end result was a program combining flexibility with good performance, which was also reflected by good scores in a user-acceptance test, with participants from the company where this project was carried out.