Aristotle University Of Thessaloniki (AUTH)
Μέθοδοι εξόρυξης γνώσης σε συλλογές περιεχομένου μεγάλης κλίμακας από κοινωνικά δίκτυα
Abstract
dc:descriptionThe recent technological developments have led to an explosive growth in the availability of online content through social networks. The massive participation of people in the content creation process has transformed the Web into a reflection of real-world entities and events. Consequently, social content analysis offers an unprecedented opportunity for knowledge extraction with respect to real-world phenomena and events. However, such an endeavor is hampered by severe challenges stemming from the attributes of social content, e.g. its variance, the lack of systematic structure and conventions, its massive scale, the prevalence of noise, and its dynamic nature. To this end, this dissertation aims at contributing to the research addressing the aforementioned challenges through modeling and studying the Web as a complex system and employing statistical and graph-based methods for achieving large-scale knowledge extraction from massive amounts of online content shared through social networks. More concretely, the dissertation investigates the following: (a) statistical properties of social content popularity, (b) survey and comparative analysis of scalability attributes of community detection methods, (c) multimedia content clustering by means of community detection on cooccurrence and similarity graphs, (d) detection and online presentation of landmarks and events in massive amounts of online social content, (e) image content understanding by integrating the structure of similarity graphs in a semi-supervised learning framework.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2012
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Papadopoulos, Symeon
- Παπαδόπουλος, Συμεών
Subjects
dc:subject × 16- Εξόρυξη γνώσης
- Ομαδοποίηση περιεχομένου
- Ανίχνευση κοινοτήτων
- Κοινωνικά δίκτυα
- Ημι-επιβλεπόμενη μάθηση
- Συστήματα κοινωνικής επισημείωσης
- Knowledge extraction
- Content clustering
- Community detection
- Social networks
- Semi-supervised learning
- Social tagging systems
- Natural Sciences
- Computer and Information Sciences
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/26668