No abstract

In medical organizations, healthcare providers need to have fast access to patients' medical information in order to make accurate diagnoses as well as to provide appropriate treatments. Efficient healthcare is thus highly dependent on doctors being provided with access to patients' medical information at the right time and place. However it frequently happens that critical pieces of pertinent information end up not being used because they are located in information systems that do not inter-operate in a timely manner. Unfortunately the standard operational mode for many healthcare applications, and even healthcare institutions, is to be managed and operated as isolated islands that do not share information in an efficient manner. There are many reasons that contribute to this grim state of affairs, but what interests us the most is the lack of enforceable security policies for systems interoperability and data exchange and the existence of many heterogeneous legacy systems that are almost impossible to directly include into any reasonable secure interoperable workflow. In this paper we propose a RBAC mobile agent access control model supported by a specially managed public key infrastructure for mobile agent's strong authentication and access control. Our aim is to create the right means for doctors to be provided with timely accurate information, which would be otherwise inaccessible, by the means of strongly authenticated mobile agents capable of securely bridging otherwise isolated institutional eHealth domains and legacy applications.

Tampered multimedia content is being increasingly used in a broad range of cybercrime activities. The spread of fake news, misinformation, digital kidnapping, and ransomware-related crimes are amongst the most recurrent crimes in which manipulated digital photos and videos are the perpetrating and disseminating medium. Criminal investigation has been challenged in applying machine learning techniques to automatically distinguish between fake and genuine seized photos and videos. Despite the pertinent need for manual validation, easy-to-use platforms for digital forensics are essential to automate and facilitate the detection of tampered content and to help criminal investigators with their work. This paper presents a machine learning Support Vector Machines (SVM) based method to distinguish between genuine and fake multimedia files, namely digital photos and videos, which may indicate the presence of deepfake content. The method was implemented in Python and integrated as new modules in the widely used digital forensics application Autopsy. The implemented approach extracts a set of simple features resulting from the application of a Discrete Fourier Transform (DFT) to digital photos and video frames. The model was evaluated with a large dataset of classified multimedia files containing both legitimate and fake photos and frames extracted from videos. Regarding deepfake detection in videos, the Celeb-DFv1 dataset was used, featuring 590 original videos collected from YouTube, and covering different subjects. The results obtained with the 5-fold cross-validation outperformed those SVM-based methods documented in the literature, by achieving an average F1-score of 99.53%, 79.55%, and 89.10%, respectively for photos, videos, and a mixture of both types of content. A benchmark with state-of-the-art methods was also done, by comparing the proposed SVM method with deep learning approaches, namely Convolutional Neural Networks (CNN). Despite CNN having outperformed the proposed DFT-SVM compound method, the competitiveness of the results attained by DFT-SVM and the substantially reduced processing time make it appropriate to be implemented and embedded into Autopsy modules, by predicting the level of fakeness calculated for each analyzed multimedia file.

Abstract. Personal mobile devices with real practical computational power and Internet connectivity are currently widespread throughout all levels of society. This is so much so that the most popular of these devices, the smart phone, in all its varied ubiquitous manifestations is nowadays the de facto personal mobile computing platform, be it for civil or even military applications. In parallel with these developments, Internet application providers like Google and Facebook are developing and deploying an ever increasing set of personal services that are being aggregated and structured over personal user accounts were an ever increasing set of personal private sensitive attributes is being massively aggregated. In this paper we describe OFELIA (Open Federated Environment for Leveraging of Identity and Authorization), a framework for user centric identity management that provides an identity/authorization versatile infrastructure that does not depend upon the massive aggregation of users identity attributes to offer a versatile set of identity services. In OFELIA personal attributes are distributed among and protected by several otherwise unrelated AAs (Attribute Authorities). Only the user mobile device knows how to aggregate these scattered AAs identity attributes back into some useful identifiable entity identity. Moreover by recurring to an IdB (Identity Broker), acting as a privacy enhancing blind caching-proxy, in OFELIA the identity attributes location in the Internet is hidden from the RP/SP (Relying Party, Service Provider) that wants to have temporary access to the users personal data. The mobile device thus becomes the means by which the user can asynchronously exercise discretionary access control over their most sensitive dynamic identity attributes in a simple but highly transparent way.