LaSID – Laboratory of Smart and Distributed Systems

Lines of Research:

Clustering

The objective of this line of research is to investigate miscellaneous aspects concerning clustering, promoting the development of new techniques and performing the application of these techniques into diverse segments for exploratory data analysis and discovery of new and relevant knowledge.

Machine Learning

In this line of research, we investigate diverse techniques of Machine Learning, evaluating their performance, proposing modifications and employing them on real problems.

Software architectures for replication

This line of research exploits algorithms, mechanisms, strategies and technologies utilized for data replication, with the purpose of harnessing the great processing power of computing clusters. As result, we produced the middleware of Rejoinder replication.

Distributed Computing

Investigate the principle of distributed algorithms and their application into diverse segments. We also investigate solutions for the problem of task scheduling, load balancing and fault tolerance inside distributed environments.

Data Mining

This line of research investigates data mining techniques and applications for the solution of real problems.

Reliable cluster replication

The objective of this research is to investigate issues concerning the construction of a full system for highly reliable data administration on clusters.

Security on Distributed Systems

In this line of research, network security models and mechanisms are researched, focused on multi-domain administrative environments, encompassing interaction between personal networks and social networks.

Decision Support Systems

In this line of research, we pursuit solutions to create a flexible architecture for unified analytical processing of structured and unstructured data, which enable to seize different interpretations and properties of unstructured data, alongside with the properties that may be stored inside structured data. To satisfy this objective, we propose the following specific objectives: propose a unified data model, which offers to capture and exploit characteristics inherent to unstructured data; offer efficient and scalable analytical processing for the proposed data model; and propose an integration model of data sources and intelligent preprocessing, as parts of the process of Extraction, Transformation and Loading, to load and update the data in the Data Warehouse. The researches involve the construction of a multidimensional data model, the analytical process and access and visualization to structured and unstructured data, and finally, validation of the proposals, exhibiting aspects concerning development and manipulation ease as the main interest.

Faculty Members:

• Prof. Dr. Gustavo Maciel Dias Vieira
• Prof. Dr. Katti Faceli
• Prof. Dr. Sahudy Montenegro González
• Prof. Dr. Tiago Agostinho de Almeida
• Prof. Dr. Tiemi Christine Sakata
• Prof. Dr. Yeda Regina Venturini

For more information concerning the group, access here.

Entry at the Directory of Research Groups in Brazil (Diretório dos Grupos de Pesquisa no Brasil) – CNPq
dgp.cnpq.br/dgp/espelhogrupo/7551185804538152