Cookies Policy
The website need some cookies and similar means to function. If you permit us, we will use those means to collect data on your visits for aggregated statistics to improve our service. Find out More
Accept Reject
  • Menu
    Article

    LIAAD coordinates project on the Extraction of Knowledge in Distributed Flows of Data

    The Laboratory of Artificial Intelligence and Decision Support (LIAAD) is currently involved in KDUS – Knowledge Discovery from Ubiquitous Data Streams, a research project financed by FCT for 2009-2012 in the scientific areas of Data Analysis and Knowledge Extraction. The main goal with this project is to study, analyse, develop and evaluate distributed and adaptive learning algorithms, which learn from continuous flows of data generated in dynamic environments.

    27th December 2009

    Two recent developments are changing the way we are in the world: on one hand, there are artefacts equipped with computing abilities, objects provided with sensors that are no longer static and inanimate and become adaptive and reactive; on the other hand, there is an explosion of all kinds of communication networks, thus making it possible to share information and to carry out self-organisation. The combination of these two technologies makes it possible to develop communities of adaptive smart devices.

    The equipments are extremely important in a sense that they are capable of sensing the environment, of receiving information from other sensors and continuously adapting to environmental changes and to the evolution of the users’ habits and needs. At the same time, with these equipments, it is possible to work with limited resources caused by computational power, memory, battery and communication restrictions. At the same time, they are capable of carrying out self-diagnosis, a characteristic of intelligence, mainly because they can foresee the possibilities of failure.

    This project includes two dimensions that should be highlighted: on one hand, the studied phenomena generate flows of data (electric distribution grids, environmental sensor networks, Websites, etc.) and, on the other hand, the information is generated in a distributed way, and thus sharing that information entails computational costs.

    The project is designed in two levels: in the first level, located in the layer of distributed data that evolve with time, LIAAD will study the methods and techniques for local model learning that evolve throughout time, and detect changes in the data generating process, auto-adapting to the most recent data. On the other hand, the second level is located on the model layer. Here, LIAAD’s team will study techniques and methods for the construction of global models from local models, modelling the evolution of the models and making self-diagnoses.

    The research team includes Pedro Rodrigues, Raquel Sebastião and Elena Ikonomovska. João Gama is LIAAD’s researcher responsible for the project.

    BIP, November 2009