Knowledge discovery for the linguistic atlas of sicily project

摘要

The identification of new useful patterns in data is a core process for intelligent systems. In this work a knowledge discovery methodology to retrieve useful and novel information from data stored in a DBMS is presented. This methodology has been designed to be used in the ALS (Linguistic Atlas of Sicily) Project. The theoretical part of the ALS methodology has been developed jointly from researchers at the Dipartimento di Scienze Filologiche e Linguistiche of the University of Palermo. The main purposes of the project are to track and to study the geo-linguistic and lexicographic processes about the function and usage of the Sicilian dialect.rnThe whole framework is based on the integration of several modules responding to different purposes. The knowledge discovery modules uses an on ontology-based methodology for the creation, retrieval, manipulation and browsing of information. The on-going collaboration process is a perfect example of a domain hybridizing process, enabling the training on-the-field of a group of researchers who participate to the constitution of the a local humanities computing community.rnThis work presents the process of knowledge discovery. Starting from conceptualization of few basic ideas, concepts have been extracted from the system's DBMS through an XML-based mapping and have been used as building blocks for further investigations. The results are incrementally and automatically proposed to researchers, who may determine to use them as new knowledge for further utilizations or can discard them.