Invited Speaker

Unmasking the Shadows: A Dive into Cutting-Edge Strategies against Disinformation in Online Social Networks

Prof. David Camacho
Universidad Politécnica de Madrid (UPM), Spain
Director of Applied Intelligence and Data Analysis research group


Abstract:
The spread of disinformation and misleading information in online social networks (OSN) has become a significant issue, causing harm at individual, community, and national levels. The rapid, and uncontrolled, dissemination through online social networks has created an environment for the proliferation of falsehoods, rumors, propaganda, and deceit, impacting the economy, politics, and health in recent decades. Addressing this challenge requires a comprehensive, multimodal approach involving various stakeholders such as individuals, media organizations, governments, technology companies, and researchers. This keynote aims to provide insights into the associated challenges and explore how Artificial Intelligence, Machine Learning and Social Network Analysis techniques are being used to combat disinformation. The keynote will specifically focus on two main areas: NLP and Multimodal DL architectures. The natural language processing, under the fact-check architecture (a solution utilizing ensembles and deep learning based on Transformers technology) will be presented to show how NLP and SNA can be effectively integrated to detect both misleading content and how it is spreading through the OSN. On the other hand, new deep learning models, designed to analysis and fusion multimodal (images and video) information, will be presented. This second architecture is used to analyse posts, which is a challenging task, due to the diverse information modalities, including text and images. To address this, a proposed multimodal architecture is introduced for identifying manipulated posts. The model is trained on a large-scale multimodal dataset encompassing image, caption, comments, and metadata. A coding system for comments is proposed to capture both semantics and tree structure. Utilizing a deep learning early fusion technique with CLIP as a pretrained encoder, hidden representations are combined based on the information channel, processing both multimodal and unimodal representations. Finally, some future trends and challenges related to the problem of detecting and combating misinformation will be presented.

Biography:
David Camacho is Full Professor at Computer Systems Engineering Department of Universidad Politécnica de Madrid (UPM), he is the head of the Applied Intelligence and Data Analysis research group, the Director of the PhD program in Computer Science and Technologies of Smart Cities, and the Director of the Master program in Machine Learning and Big Data at UPM. He has published more than 300 journals, books, and conference papers. His research interests include Machine Learning (Clustering/Deep Learning), Computational Intelligence (Evolutionary Computation, Swarm Intelligence), Social Network Analysis, Fake News and Disinformation Analysis. He has participated/led more than 60 AI-based R&D projects (National and International: H2020, MCSA ITN-ETN, DG Justice, ISFP, NRF Korea), applied to real-world problems in areas as aeronautics, aerospace engineering, cybercrime/cyber intelligence, social networks applications, disinformation countering, or video games among others. He serves as Editor in Chief of Expert Systems from 2023 and sits on the Editorial Board of several journals including Information Fusion, Human-centric Computing and Information Sciences (HCIS), and Cognitive Computation, IEEE Transactions on Emerging Topics in Computational Intelligence (IEEE TETCI), among others.

Reference:
Contact at: David.Camacho@upm.es
AIDA: https://aida.etsisi.uam.es
Google Scholar: https://scholar.google.com/citations?hl=es&user=fpf6EDAAAAAJ#
ResearchGate: https://www.researchgate.net/profile/David-Camacho-12