Computer Systems

Representative

Description

Historically, interest in the area of computing in the CSP began in the 1970s, with the development of numerical methods for analyzing structures. With increasingly complex problems involving, for example, soil-fluid-structure interactions, and with the uncertainties inherent in natural phenomena, research in the field of structures broadened its scope, which fostered the need for other computational resources. In the 1980s, the emergence of parallel and vector computers made it possible to solve more complex problems involving a large number of variables. Computational methods, previously restricted to deterministic solutions, began to deal with uncertainties. From the 1990s onwards, the evolution of algorithms and the increase in computer capacity allowed robust models to be developed from large databases. In recent years, research has focused on extracting knowledge from large collections of documents.

Research in the area of Computational Systems focuses on the development of computational models, using computational resources in parallel and distributed environments, databases, graphic visualizations, etc. The main focus of research has been on the development of Computational Intelligence and Data Mining techniques in the modeling of complex systems. The work involves the development of algorithms based on fuzzy logic, neural networks, support vector machines, genetic algorithms and optimization techniques inspired by nature. These techniques have diverse applications in practically all areas of knowledge, which allows the development of projects in conjunction with the other areas of concentration in the PEC and other COPPE programs.

The work has been carried out in institutional and technological research projects, focused on developing models of complex systems in a variety of applications in the oil industry, the environment, energy, business and others. The applications are not limited to the field of Engineering and the area also works on problems in bioinformatics, remote sensing, ecology, among others, always in cooperation with specialists in these areas.

The impact of the Internet on human development has been compared to the invention of the printing press in terms of the profound technological, economic and social transformations brought about by the dissemination of information on the “big web”. In this context, the development of scientific research in the area of computer systems has been oriented towards the extraction of knowledge from collections of documents and methodologies to enhance the generation of knowledge and cooperative networking.

Several research projects are carried out in collaboration with other COPPE programs. The Technology Transfer Center (NTT) is the PEC laboratory that works specifically in this area, and has a strong interaction with the Laboratory of Computational Methods in Engineering (LAMCE) and COPPE’s High Performance Computing Center (NACAD).

The figure above shows the analysis of the theses and dissertations of the Civil Engineering Program, showing 7 different areas. In the graph, each node represents a work and the length of the edge between two works is proportional to the similarity.

Lines of Research

This line of research investigates the various aspects of complexity in technological, biological and social systems. It addresses cognitive processes for learning certain systems and solving problems. It uses fractals in physical phenomena, without restricting itself to geometric representations. It develops knowledge transfer models. Research projects aim to integrate different technologies.

Development of optimization algorithms inspired by nature, such as genetic algorithms, swarm intelligence, artificial immune systems, differential evolution, among others. Research projects in this line are aimed at improving algorithms and applications in various areas, including not only complex engineering problems (such as systems for offshore oil production), but also interdisciplinary applications including, for example, computational biology problems (such as protein folding, computational molecular docking and peptide sequencing from mass spectrometry).

Development of new computational intelligence algorithms for data mining in various applications. This line of research deals mainly with data models that can be useful for extracting knowledge in complex engineering, bioinformatics and business applications, among others.

Development of algorithms and systems for text and web mining. Research projects in this line address all stages of the text mining process: pre-processing, adaptation of data mining algorithms for text mining applications, visualization, knowledge discovery on the Internet (web mining: navigation, content and link analysis), among others.

This line of research aims to aggregate and process large masses of data obtained in real time from various repositories and web-plugged sensors that continuously capture structured and unstructured information in order to generate knowledge and meet the criteria of efficiency and privacy. It is of interest in various fields and activities that need to analyze information from large volumes of data, such as: oil and gas, biology, the environment, meteorology, satellite images, social networks, competitive intelligence, security, the financial market, e-commerce, etc.

This line of research investigates methods for analyzing uncertainties and risks in various applications. The research projects in this line are aimed at developing new methodologies for analyzing uncertainties in applications in the financial sector, oil and ecology and decision-making processes.

It covers the theory of complex networks and the mathematical formalism of graph theory. Development of new computational intelligence algorithms for various applications. Research projects aim to integrate complex systems modeling technologies.

This line of research uses cell phone data in which the record of each call is associated with the geographical coordinates of the antenna that processed the call. Studies show that human movement has a high degree of temporal and spatial regularity. After correcting for differences in distance and the anisotropy inherent in each trajectory, the displacement patterns converge on the same probability distribution, indicating that, despite the great diversity of trajectories, human mobility follows reproducible patterns. In this way, efficient models of urban dynamics can be built for areas such as transportation, the spread of viruses, epidemics, land use and so on.

Development of systems for visualizing the results of numerical modeling and data mining processes.