Abstract
dc:descriptionDue to the recent "Open Science'' initiatives, vast amounts of scholarly data (e.g., publications, authors, venues, topics) have become publicly available. These data, being inherently heterogeneous, are naturally represented in the form of scholarly information networks. This graph representation, being fairly structured, yet flexible, enables mining and analysing scholarly knowledge from multiple perspectives. In this thesis, we exploit such semantically rich data mining and analysis techniques for scholarly information networks, that reveal latent patterns (and relationships) encoded in the structure of the network, focusing on three main directions, namely (i) impact assessment of scientific publications, (ii) expert finding in academia, and (iii) evaluation of the underlying queries. We propose methods focused on specific problems in the aforementioned directions, aiming to advance the relevant state-of-the-art. Our results showcase the efficiency and effectiveness of our methods, highlighting that latent knowledge extracted from scholarly information networks is invaluable in a variety of practical problems and applications.
Degree
thesis:*- Grantor dc:publisher
- University of Peloponesse
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Chatzopoulos, Serafim
- Χατζόπουλος, Σεραφείμ
Subjects
dc:subject × 16- Εξόρυξη δεδομένων
- Βιβλιογραφικά δίκτυα πληροφορίας
- Εκτίμηση απήχησης ερευνητικών δημοσιεύσεων
- Αποτίμηση ερωτημάτων
- Ετερογενή δίκτυα πληροφορίας
- Data mining
- Scholarly information networks
- Publications impact assessment
- Query evaluation
- Heterogeneous information networks
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Υπολογιστές, Υλικό (hardware) και Αρχιτεκτονική
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Computer science, Hardware and Architecture
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/52103