ECO Research Project

Universidade do Minho

This project is in the area of Extraction of Knowledge (Knowledge Discovery in Data Bases or Data Mining), which is rather recent but expanding rapidly. The objective of the work is to find methods for non-trivial extraction of information which is implicit in the existing data. The goal is to obtain information which is potentially useful and which can be represented in a high-level language so as to facilitate interpretation.
As the aim is to apply the methods to real data, that is, volumes of data which tend to be relatively large, the work involves adaptation of some data pre-processing techniques whose objective is to reduce its complexity. The second objective involves the adaptation of AI techniques, and particularly those of machine learning.

ECO Partners

  • Universidade do Porto, LIACC
  • Universidade Nova de Lisboa, Depart. Informatica
  • Universidade Nova de Lisboa, Depart. Matematica
  • Instituto Nacional de Estatistica
  • Universidade do Minho, Depart. Informatica

    Our group is particularly concerned in applying deductive databases techniques to improve the ILP algorithms, namely query-answering techniques. Also, we intend to investigate novel techniques for the induction of databases attribute dependencies (and other data dependencies) from relational databases using novel ideas of subsumption from a logic programming perspective. This is also related to the desirable features of extracting knowledge from databases in constant mutation.

    ECO Related Papers

  • Magic Sets with Full Sharing (Technical Report).
  • Magic Sets with Full Sharing (to appear at the Journal of Logic Programming).
  • Tuple Tabulation (Technical Report).

    mantained by Paulo Azevedo