A Decentralized and Robust Approach to Estimating a Probabilistic Mixture Model for Structuring Distributed Data

Abstract

HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Computing Privacy Risk and Trustworthiness of Users in SNSs

“…Additionally it has been found, that social relationships have significant influence on the trust, so type of accepted friend requests (e.g. friends, strangers or known in person) will help to protect privacy [9]. We define customized privacy settings as all the personal information that are hidden or shared with selected people, by users preferences.…”

“…This proof social sites to be attack vectors with different functionalities such as posts, friend requests, photos and videos uploads, likes, comments, third party applications, advertisements and links to other websites making them potential avenues of information leakage [5], [14]. Privacy risk on disclosure of individual's sensitive information through online social interactions can be protected through access control [6] and algorithms like anonymization [8], [12] and decentralized modeling [9]. This represents a social network as a graph, applying various transformations to achieve security.…”

Large-scale unsupervised spatio-temporal semantic analysis of vast regions from satellite images sequences

“…CNNs are able to extract patterns from massive volumes of complex data, and they become a natural candidate to tackle problems in the field of remote sensing. The use of deep embeddings based on CNNs is gaining increasing attention due to their ability to encode the images into relevant latent features used later in other classification or clustering methods (Ji et al 2018;Kalinicheva et al 2020;Attar et al 2011). However, the feature extraction obtained by the embedding is usually guided by a good compression of the information contained in the original image, and it is not directly related to a classification or clustering purpose.…”