SPBDIoT 2018 Abstracts


Full Papers
Paper Nr: 1
Title:

Context-aware and Attribute-based Access Control Applying Proactive Computing to IoT System

Authors:

Noé Picard, Jean-Noël Colin and Denis Zampunieris

Abstract: ABAC allows for high flexibility in access control over a system through the definition of policies based on attribute values. In the context of an IoT-based system, these data can be supplied through its sensors connected to the real world, allowing for context-awareness. However, the ABAC model alone does not include proposals for implementing security policies based on verified and/or meaningful values rather than on raw data flowing from the sensors. Nor does it allow to implement immediate action on the system when some security flaw is detected, while this possibility technically exists if the system is equipped with actuators next to its sensors. We show how to circumvent these limitations by adding a proactive engine to the ABAC components, that runs rule-based scenarios devoted to sensor data pre-processing, to higher-level information storage in the PIP, and to real-time, automatic reaction on the system through its actuators when required.

Paper Nr: 4
Title:

Homomorphic Encryption for Secure Computation on Big Data

Authors:

Roger A. Hallman, Mamadou H. Diallo, Michael A. August and Christopher T. Graves

Abstract: With the ubiquity of mobile devices and the emergence of Internet of Things (IoT) technologies, most of our activities contribute to ever-growing data sets which are used for big data analytics for a variety of uses, from targeted advertising to making medical and financial judgments and beyond. Many individuals and organizations adopt this new big data paradigm without giving any consideration to privacy and security when they create this data and voluntarily give it up for aggregation. Data breaches have become such a common occurrence that it is easy to despair that concepts like privacy and security are antiquated and we should simply accept data leakage as a new normal. Homomorphic Encryption (HE) is a method of secure computation which allows for calculations to be made on encrypted data without decrypting it and without giving away information about the operations being done. While HE has historically been plagued by computational inefficiencies, the field is rapidly advancing to a point where it is efficient enough for practical use in limited settings. In this paper, we argue that, with sufficient investment, HE will become a practical tool for secure processing of big data sets.

Area 1 - Software Agents and Internet Computing

Full Papers
Paper Nr: 5
Title:

A Discussion Paper on the Grey Area – The Ethical Problems Related to Big Data Credit Reporting

Authors:

Victor Chang and Jing Li

Abstract: With the rise and the development of the “credit society”, the credit reporting has played a central role in evaluation one’s credit statues, including monitoring and updating creditworthiness of individuals. As the emergence of big data, new tools enabling the credit reporting system to develop new level, by collecting the online and offline data to establish more completely score system. This review paper is aimed to present the difference between the new big data credit reporting and traditional credit reporting, and then explain advantages offered by the new data management. Subsequently, ethical problems will be described due to rising concerns. Being “kidnapped” by the credit reporting applications, users’ data will be collected and disposed without prior permission. Some data processes may arise with the messy, unreasonable and fake data resource problems to add more complexities to the existing services which are unable to cope with. As a result, individual users could not verify the correctness of the data and did not know which data would be more trustworthy to be verified for payment and billing. To be worse, users even do not know how to improve their creditworthiness if they have done everything correctly. There are some issues about precision marketing, since some data brokers will target the individuals who was vulnerable to the non-performing and short-term loans. Last but not least, the algorithm of big data prone to evaluate the credit score by groups that individual related to, rather than the individual’s own merits, which may lead to discrimination issue, and accelerate the wealth gap problem.