La Gatta, Valerio (2023) Knowledge-informed Disinformation Mining: From Fact-Checking to Content Moderation. [Tesi di dottorato]
Anteprima |
Testo
LaGatta_Valerio_36.pdf Download (10MB) | Anteprima |
| Tipologia del documento: | Tesi di dottorato |
|---|---|
| Lingua: | English |
| Titolo: | Knowledge-informed Disinformation Mining: From Fact-Checking to Content Moderation |
| Autori: | Autore Email La Gatta, Valerio valerio.lagatta@unina.it |
| Data: | 12 Dicembre 2023 |
| Numero di pagine: | 212 |
| Istituzione: | Università degli Studi di Napoli Federico II |
| Dipartimento: | Ingegneria Elettrica e delle Tecnologie dell'Informazione |
| Dottorato: | Information technology and electrical engineering |
| Ciclo di dottorato: | 36 |
| Coordinatore del Corso di dottorato: | nome email Russo, Stefano stefano.russo@unina.it |
| Tutor: | nome email Moscato, Vincenzo [non definito] |
| Data: | 12 Dicembre 2023 |
| Numero di pagine: | 212 |
| Parole chiave: | disinformation mining, content moderation, fact-checking, knowledge-informed methods |
| Settori scientifico-disciplinari del MIUR: | Area 09 - Ingegneria industriale e dell'informazione > ING-INF/05 - Sistemi di elaborazione delle informazioni |
| Depositato il: | 12 Dic 2023 10:53 |
| Ultima modifica: | 10 Mar 2026 14:23 |
| URI: | http://www.fedoa.unina.it/id/eprint/15648 |
Abstract
In the digital era, the widespread propagation of disinformation presents significant threats to societal, economic, and political stability, a concern underscored by recent global events such as the COVID-19 pandemic. This thesis adopts an integrative approach, combining computer science, network science, artificial intelligence, and knowledge-informed methodologies to tackle online disinformation. Recognizing disinformation as a complex phenomenon, intertwined with human cognition, social dynamics, and emotional responses, the research focuses on leveraging diverse forms of contextual knowledge to combat this challenge. Focusing on the enduring importance of manual fact-checking processes, we introduce an AI-driven system to expedite fact-checking by utilizing knowledge from previously fact-checked information. Additionally, the thesis also presents KERMIT (Knowledge-EmpoweRed Model In harmful meme deTection), an innovative methodology that combines internal meme content with background and cultural knowledge for harmful meme detection. Furthermore, we explore how simultaneously addressing various disinformation-related tasks, such as fake news detection and sentiment analysis, can bolster overall detection performance and provide a deeper understanding of disinformation content. Lastly, the thesis investigates how the knowledge of content moderation on a source platform can inform the moderation strategies of the other social media platforms, enhancing the integrity of the overall digital information ecosystem. All in all, this thesis advances the understanding of online disinformation and underscores the need for holistic, knowledge-driven approaches to address this pervasive issue.
Downloads
Downloads per month over past year
Actions (login required)
 |
Modifica documento |
