Athens University Economics and Business (AUEB)
Σημασιολογικός χαρακτηρισμός, οργάνωση και διαχείριση περιεχομένου του Παγκόσμιου Ιστού, με χρήση οντολόγιων και έμφαση στο ρόλο των υπερσυνδέσμων
Abstract
dc:descriptionThe dissertation describes a new method of collecting, characterizing and organizing the contents of the World Wide Web (WWW). The difference of the method from the existing ones lies in the use of a thematic ontology at all levels. The ontology describes in each case the field of interest, and therefore the documents collected and organized constitute a THEMatic SUbset of the WWW. For a thematic subset of the WWW documents, the method of organization is divided into three main stages: a) the characterization of the documents with lexical and semantic features, b) the organization of the documents in groups with common characteristics and c) the management of the collected and organized information.The term "lexical features" of a document refers to the set of words that describe it, while the term "semantic features" refers to the set of ontological concepts to which these words correspond. The organization of WWW documents in thematic subsets is based on the variety of their semantic characteristics. To calculate the quality of two documents, a new measure is introduced that takes into account the distance - in the ontology - of the sets of concepts described by the two documents. This measure is not based on the absolute lexical similarity between the two descriptions, as is the case with the existing similarity measures, but on the semantic similarity they show. For this reason, it is more flexible and gives better results.The words and meanings extracted for each document are stored in a separate XML file, which can be used by other applications but also contained in the original document as a metadata file. The structure of each XML document is strictly described by an XML-Schema file. To facilitate the processing of the collected information (queries, knowledge mining, etc.), the XML file data is stored in a relational database.In the framework of the dissertation, the THESUS language was developed, which defines a set of operators for the management of hyperlinks and the information they carry, and the THESUS system, which implements the proposed method of managing WWW documents and at the same time uses a new similarity measure for documents. Also develops a methodology for mapping XML structures, using XML-Schema, to the relational model as well as the X-Database system, which undertakes the automatic creation of relational databases from the XML-Schema and the storage, management and retrieval of XML documents in the relational database.
Degree
thesis:*- Grantor dc:publisher
- Athens University Economics and Business (AUEB)
- Year dc:date
- 2003
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Varlamis, Iraklis
- Βαρλάμης, Ηρακλής
Subjects
dc:subject × 17- Εξόρυξη γνώσης σε μεγάλη κλίμακα
- Ανάκτηση πληροφοριών από κείμενα
- Κατηγοριοποίηση κειμένων
- Γλωσσικοί θησαυροί
- Παγκόσμιος ιστός
- XML
- Big data mining
- Information retrieval
- Automatic text classification
- Linguistic thesauri
- WWW
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Πληροφοριακά συστήματα
- Computer and Information Sciences
- Natural Sciences
- Information Systems
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/49798