People

Research activity in 2000

In 2000, the group continued to be active on several topics in collaboration with the group of Machine Learning at the Università del Piemonte Orientale at Alessandria: for what concerns learning techniques in First Order Logic, algorithms previously developed in the systems NTR, FONN and G-Net have been further extended and integrated in a unique framework. Moreover, a boosting technique has been devised and integrated in the same learning framework.

The studies on Phase Transitions in Matching and Learning, started in 1998, continued also in 2000 with renewed strength based on the results of the analysis we performed on previous experimentation: in particular, we investigated the possibility of combining stochastic sampling of the solution space with hill-climbing search strategy. Preliminary results of these studies have been published in two conferences (see references section).

The research activity of the group continues also around the extraction of statistical dependencies among data. The new paradigm of data dependence, that allows to find the actual significant dependencies among data, is studied in depth and finds some applications not only for structured data and databases but also for WEB data and texts (see references section).

Furthermore, the concept of data dependence has been studied from the viewpoint of the Information Theory, and more in particular, a renewed definition of the mutual information is investigated. Finally, the concept of support of patterns (such as the itemsets) that in Data Mining is based on simple statistical measures is currently analysed in order to propose a well-founded, theoretical definition based on the Bayes' Theorem.

A new branch of research, centered on the analysis of the role that databases play in the knowledge discovery process, has been started in 2000. This new area of research named "inductive databases" aims at studying the potentialities of query languages in the iterative process of knowledge discovery. Inductive databases are studied also from the viewpoint of the evaluation and execution of the queries.

For what concerns national projects, Marco Botta has been involved in a collaboration with CSELT (the research labs of Telecom Italia) aimed at developing a tool for pruning and visualize association rules. Moreover, the group is involved in the national project 'Intelligent Agents: Interaction and Knowledge Acquisition', whose objective is the realization of a multi-agent architecture based technology, that allows one to build applications for the extraction and synthesis of knowledge from structured and unstructured data accessible on the Web.