University of Thessaly (UTH)
Semantic social modeling of online user - generated content
Abstract
dc:descriptionIn this thesis, we study information management issues that arise in Online Social Networks (OSNs), as well as collective intelligence issues towards automated knowledge representation. We focus on three research directions, namely i) online social influence and the discovery of its impactful entities, ii) user-generated content and the role of semantics in social analysis, and iii) qualitative assessment of viral disseminated content. We present efficient and scalable methods focused on specific problems in the addressed directions, while we aim at proposing advancements in the relevant state of the art research in the field. In particular, in the first research direction, we study how we can measure social influence and what are its application domains. To this end, we developed a service aiming at calculating and ranking the importance and influence of Twitter accounts. This service incorporates theoretical aspects of influence metrics that derive from social functions that evaluate i) the activity of a Twitter account (e.g. tweets, re-tweets and replies), ii) its social degree (e.g. followers and following) and iii) its network impact (e.g. content diffusion and social acknowledgement). In the second research direction, we investigate the role of semantics in OSNs and the adoption of Semantic Web technologies which can be used for the detection of similar users, as well as user personalization issues (e.g. interests and suggestions). To this end, we define an ontological schema towards the semantification of social analytics, including structural aspects of Twitter accounts, disseminated entities and social relationships. Furthermore, we propose a methodology towards the discovery and suggestion of similar Twitter accounts, based entirely on their disseminated content. On top of that and based on the similarity relationships, we present an approach towards the automatic labeling of Twitter accounts by exploiting information from the Linked Open Data cloud; specifically according to DBpedia thematic categories. Finally, we contribute in the field of Query Expansion (QE) by proposing an algorithmic approach, which expands a user’s query through the creation of a suggestion set that consists of the most viral and up-to-date Twitter entities. Finally, in the third research direction, we tackle the problem of qualitative assessment of user-generated content by utilizing social influence and semantics. We conclude that the first two research areas along with the later can jointly provide useful insights, when we want to model dynamic properties of influential content and its flow dynamics.
Degree
thesis:*- Grantor dc:publisher
- University of Thessaly (UTH)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Razis, Gerasimos
- Ραζής, Γεράσιμος
Subjects
dc:subject × 20- Σημασιολογικός ιστός
- Ανάλυση κοινωνικών δικτύων
- Οντολογία
- Κοινωνική επιρροή
- Ομοιότητα
- Ανοιχτά συνδεδεμένα δεδομένα
- Επέκταση ερωτήματος
- Κοινωνική ποιότητα
- Semantic web
- Social network analysis
- Ontology
- Social influence
- Similarity
- Linked open data
- Query expansion
- Social quality
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Natural Sciences
- Computer and Information Sciences
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/45542