Software being developed today can have years worth of history and hundreds if notthousands of files involved in a single project. When trying to determine what parts ofthe code need maintenance or updating it can be difficult to determine what will beproblematic in the future. Hours spent on code that will not cause problems in thefuture could be better used in other areas. Before changes are made to a codebase, themost error-prone parts of the code should be identified. In this thesis a method forcomparing what factors contribute to future bugs is developed, as well as testing severalfactors extracted from version control metadata using this method. In addition, avisualization was made using tree maps to show the most problematic files in a readablemanner, effectively using the produced data in an application to predict future bugs. Itwas determined that Age of newest change, Changes with age reducing importance andPrevious bugfixes with age reducing importance were all the most impactful factors forpredicting future bugs but that different repositories worked best with differentcombinations of the mentioned factors.

Time-dependently changing hierarchical data structures are commonly visualized with animated treemaps. Changes in the data structure produce correlated changes in the animation. While previous treemap animations exclusively consider hierarchical data where only the size of the data points are time-dependently changing, we present a novel method for animating treemaps where the hierarchy of the data structure is also evolving. A change in the hierarchy is referred to as a topological event of the animated treemap. Furthermore, animation stability is a quantitative indicator as to the visual quality of treemap animations. With the proposed method, the stability of animated Slice-and-Dice treemaps is maximized during topological events. Thus, the method produces topologically evolving Slice-and-Dice treemap animations of high visual quality. Inspiration for the proposed method was largely drawn from Köpp and Weinkauf’s recent work on the ordered visualization of time-dependently evolving nested graphs. The research question regarded whether their treemap method for spatially ordering hierarchical data could be translated to the 2D treemap space, in order to optimize the stability of animated Slice-and-Dice layouts. This study does not only report on the development of the proposed method, but also present an accompanying experimental evaluation. While evidence show that the proposed method is incapable of entirely mitigating decreasing stability scores, caused by any possible topological event in an evolving Slice-and-Dice hierarchy, the success of the proposed method has been proven for the most frequent topological events. / Tidsberoende förändringar i hierarkiska datastrukturer visualiseras ofta med animerade trädkartor (efter engelskans treemap). Förändringar i datastrukturen producerar korrelerade förändringar i animationen. Medan tidigare trädkartsanimationer sett till hierarkisk data där endast storleken på datapunkterna förändras över tid, så presenterar vi en ny metod för animation av trädkartor där även datastrukturens hierarki förändras. En förändring av hierarkin kallas för en topologisk händelse i den animerade trädkartan. Animationsstabilitet är vidare en kvantitativ indikator för animerade trädkartors visuella kvalitet. Med den föreslagna metoden så har stabiliteten av animerade Slice-and-Dice-trädkartor optimerats under topologiska händelser. Metoden producerar följaktligen topologiskt föränderliga Slice-and-Dice-animationer med hög visuell kvalitet. Studien har huvudsakligen inspirerats av Köpp och Weinkaufs arbete med den sorterade visualiseringen av tidsföränderliga, nästlade grafer. Forskningsfrågan gällde huruvida deras metod för att ordna den hierarkiska datan kunde översättas till trädkartors tvådimensionella rymd, för att i ett vidare steg optimera stabiliteten av animerade Slice-and-Dice-trädkartor. Studien rapporterar ej enbart om utvecklingen av den föreslagna metoden, utan presenterar även en medföljande experimentell utvärdering. Medan man ur bevisen kan utläsa att metoden ej fullständigt kan mildra minskningar i animationsstabilitet, orsakat av alla de möjliga topologiska händelser som kan ske i en föränderlig Slice-and-Dice-hierarki, så har den föreslagna metoden varit framgångsrik i de mest frekventa fallen.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / The interpretation and understanding of large quantities of data is a challenge for current information visualization methods. The visualization of information is important as it makes the appropriate acquisition of the information through the visualization possible. The choice of the most appropriate information visualization method before commencing with the resolution of a given visual problem is primordial to obtaining an efficient solution. This article has as its objective to describe an information visualization classification approach based on Treemap, which is able to identify the best information visualization model for a given problem. The implementation of the model involves choosing components and also the architecture that stands its necessary functions effectively. This is understood through the construction of an adequate information visualization meta-model. Firstly, the actual state of the visualization field is described, and then the rules and criteria used in our research are shown, with the aim of presenting a meta-model proposal based on treemap visualization methods. Besides this, the author present a case study with the information contained in the periodic table visualization meta-model along with an analysis of the information search time complexity in each of the two meta-models. Finally, an evaluation of the results is presented through the experiments conducted with users and a comparative analysis of the methods based on Treemap and the Periodic Table. / Interpretar e entender uma grande quantidade de dados é um desafio dos modelos de Visualização de Informação. A Visualização da Informação é importante para possibilitar uma apropriada obtenção da informação através da visualização. A escolha do modelo de visualização de informação mais apropriado antes de iniciar a resolução de um determinado problema visual é primordial para obtenção de uma solução eficaz. A implementação do modelo envolve escolher também componentes e a arquitetura que suporte as suas funcionalidades necessárias, de maneira eficaz. Esta dissertação descreve uma abordagem de classificação das visualizações da informação baseado em treemap, capaz de buscar e identificar o melhor modelo de visualização da informação para um determinado problema. Isso compreende na criação de um metamodelo de Visualização da Informação adequado. Primeiramente, o atual estado do campo de visualização foi descrito e então, são mostradas as regras e critérios utilizados na pesquisa dessa dissertação, a fim de apresentar uma proposta de metamodelo baseado em treemap para modelos de visualização. Além disso, é apresentado um estudo de caso com as informações contidas no metamodelo de visualização da tabela periódica e a análise da complexidade de tempo da busca da informação em cada um dos dois metamodelos. Finalmente, uma avaliação dos resultados é apresentada, através de experimentos com usuários e análise comparativa entre os modelos baseados em treemap e na tabela periódica. / Mestre em Ciências

Treemaps are a popular space-filling visualization of hierarchical data that maps an attribute of a datum, or a data aggregate, to a proportional amount of area. Assuming a rectangular treemap consisting of nested rectangles (also called tiles), there are multiple possible valid tiling arrangements. A common criterion for optimization is aspect ratio. Nevertheless, treemaps usually consist of multiple rectangles, so the aspect ratios need be aggregated. The basic definition of aspect ratio (width divided by height) cannot be meaningfully aggregated. Given this, a definition of aspect ratio that does not differentiate height from width was suggested. This definition allows for meaningful aggregation, but only as long as there are no large differences in the data distribution, and the target aspect ratio is 1:1. Originally, a target aspect ratio of 1:1 was deemed to be axiomatically ideal. Currently, perceptual studies have found an aspect ratio of 1:1 to lead to the largest area estimation error. However, with any other target this definition of aspect ratio cannot be meaningfully aggregated. This thesis suggests a correction that can be applied to the current metric and would allow it to be meaningfully aggregated even when there are large value differences in the data. Furthermore, both the uncorrected and corrected metrics can be generalized for any target (i.e. targets other than 1:1). Another issue with current evaluation techniques is that algorithm fitness is evaluated through Monte Carlo trials. In this method, synthetic data is generated and then aggregated to generate a single final result. However, tiling algorithm performance is dependant on data distribution, so a single aggregateresult cannot generalize overall performance. The alternative suggested in this thesis is visual cluster analysis, which should hold more general predictive power.All of the above is put into practice with an experiment. In the experiment, a new family of tiling algorithms, based on criteria derived from the results of the perceptual tests in literature,is compared to the most popular tiling algorithm, Squarify. The results confirm that there are indeed vast but consistent value fluctuations for different normal distributions. At least for a target aspect ratio of 1.5, the new proposed algorithms are shown to perform better than Squarify for most use cases in terms of aspect ratio.