Examinando por Autor "Zubiaga, Arkaitz"

Mostrando 1 - 2 de 2
  • Miniatura
    Ítem
    Analyzing the existence of organization specific languages on Twitter
    (Institute of Electrical and Electronics Engineers Inc., 2021-08-05) Sánchez Corcuera, Rubén; Zubiaga, Arkaitz ; Almeida, Aitor
    The presence of organisations in Online Social Networks (OSNs) has motivated malicious users to look for attack vectors, which are then used to increase the possibility of carrying out successful attacks and obtaining either private information or access to the organisation. This article hypothesised that organisations have specific languages that their members use in OSNs, which malicious users could potentially use to carry out an impersonation attack. To prove these specific languages, we propose two tasks: classifying tweets in isolation by their author's organisation and classifying users' entire timelines by organisation. To accomplish both tasks, we generate a dataset of over 15 million tweets of five organisations, and we apply language dependant models to test our hypothesis. Our results and the ablation study conclude that it is possible to classify tweets and users by organisation with more than three times the performance achieved by a traditional ML algorithm, showing a substantial potential for predicting the linguistic style of tweets.
  • Miniatura
    Ítem
    Early detection and prevention of malicious user behavior on Twitter using deep learning techniques
    (Institute of Electrical and Electronics Engineers Inc., 2024) Sánchez Corcuera, Rubén; Zubiaga, Arkaitz; Almeida, Aitor
    Organized misinformation campaigns on Twitter continue to proliferate, even as the platform acknowledges such activities through its transparency center. These deceptive initiatives significantly impact vital societal issues, including climate change, thus spurring research aimed at pinpointing and intercepting these malicious actors. Present-day algorithms for detecting bots harness an array of data drawn from user profiles, tweets, and network configurations, delivering commendable outcomes. Yet, these strategies mainly concentrate on postincident identification of malevolent users, hinging on static training datasets that categorize individuals based on historical activities. Diverging from this approach, we advocate for a forward-thinking methodology, which utilizes user data to foresee and mitigate potential threats before their realization, thereby cultivating more secure, equitable, and unbiased online communities. To this end, our proposed technique forecasts malevolent activities by tracing the projected trajectories of user embeddings before any malevolent action materializes. For validation, we employed a dynamic directed multigraph paradigm to chronicle the evolving engagements between Twitter users. When juxtaposed against the identical dataset, our technique eclipses contemporary methodologies by an impressive 40.66% in F score (F1 score) in the anticipatory identification of harmful users. Furthermore, we undertook a model evaluation exercise to gauge the efficiency of distinct system elements.