Mould and yeast identification in archival settings: preliminary results on the use of traditional methods and molecular biology options in Portuguese archives

This project was developed to fully assess the indoor air quality in archives and libraries from a fungal flora point of view. It uses classical methodologies such as traditional culture media – for the viable fungi – and modern molecular biology protocols, especially relevant to assess the non-viable fraction of the biological contaminants. Denaturing high-performance liquid chromatography (DHPLC) has emerged as an alternative to denaturing gradient gel electrophoresis (DGGE) and has already been applied to the study of a few bacterial communities. We propose the application of DHPLC to the study of fungal colonization on paper-based archive materials. This technology allows for the identification of each component of a mixture of fungi based on their genetic variation. In a highly complex mixture of microbial DNA this method can be used simply to study the population dynamics, and it also allows for sample fraction collection, which can, in many cases, be immediately sequenced, circumventing the need for cloning. Some examples of the methodological application are shown. Also applied is fragment length analysis for the study of mixed Candida samples. Both of these methods can later be applied in various fields, such as clinical and sand sample analysis. So far, the environmental analyses have been extremely useful to determine potentially pathogenic/toxinogenic fungi such as Stachybotrys sp., Aspergillus niger, Aspergillus fumigatus, and Fusarium sp. This work will hopefully lead to more accurate evaluation of environmental conditions for both human health and the preservation of documents.

General Ribeiro Arthur militar e artista

Ribeiro Arthur foi militar, crítico de Arte e foi sobretudo um dos mais importantes investigadores e ilustradores da história dos uniformes em Portugal. Produziu um número muito significativo de aguarelas dedicadas a este tema que, reunidas em colecção, estão preservadas no Arquivo Histórico Militar. ABSTRACT - Artur Ribeiro Arthur was military, art critic and was one of the most important researchers and illustrators in the history of the uniforms in Portugal. He produced a significant number of watercolors dedicated to this theme, gathered in collections that are kept in the Military Historical Archive.

Risk analysis in a Portuguese archive what has changed in five years?

The Cultural Property Risk Analysis Model was applied in 2006 to a Portuguese archive located in Lisbon. Its results highlighted the need for the institution to take care of risks related to fire, physical forces and relative humidity problems. Five years after this first analysis the results are revisited and a few changes are introduced due to recent events: fire and high humidity remain an important hazard but are now accompanied by a pressing contaminants problem. Improvements in storage systems were responsible for a large decrease in terms of calculated risk magnitude and proved to be very cost-effective.

Leonel Duarte Ferreira (1894-1959) : vida e obra musical

Tese apresentada para o cumprimento dos requisitos necessários à obtenção do grau de Doutor no ramo de Ciências Musicais

Segmentação automática de memórias pessoais

Dissertação para obtenção do grau de Mestre em Engenharia Informática e de Computadores

Text classification using compression-based dissimilarity measures

Arguably, the most difficult task in text classification is to choose an appropriate set of features that allows machine learning algorithms to provide accurate classification. Most state-of-the-art techniques for this task involve careful feature engineering and a pre-processing stage, which may be too expensive in the emerging context of massive collections of electronic texts. In this paper, we propose efficient methods for text classification based on information-theoretic dissimilarity measures, which are used to define dissimilarity-based representations. These methods dispense with any feature design or engineering, by mapping texts into a feature space using universal dissimilarity measures; in this space, classical classifiers (e.g. nearest neighbor or support vector machines) can then be used. The reported experimental evaluation of the proposed methods, on sentiment polarity analysis and authorship attribution problems, reveals that it approximates, sometimes even outperforms previous state-of-the-art techniques, despite being much simpler, in the sense that they do not require any text pre-processing or feature engineering.