Inference attacks based on neural networks in social networks

2017 ◽  
Author(s):  
Bo Mei ◽  
Yinhao Xiao ◽  
Hong Li ◽  
Xiuzhen Cheng ◽  
Yunchuan Sun
Keyword(s):  
Neural Networks ◽  
Social Networks ◽  
Inference Attacks

scholarly journals Social Security and Privacy for Social IoT Polymorphic Value Set: A Solution to Inference Attacks on Social Networks

2019 ◽  
Vol 2019 ◽  
pp. 1-16
Author(s):  
Yunpeng Gao ◽  
Nan Zhang
Keyword(s):  
Social Networks ◽  
Social Network ◽  
Internet Of Things ◽  
Wide Distribution ◽  
Security And Privacy ◽  
Ranking Functions ◽  
Retrieval Models ◽  
Utility Loss ◽  
Inference Attacks ◽  
Real World Datasets

Social Internet of Things (SIoT) integrates social network schemes into Internet of Things (IoT), which provides opportunities for IoT objects to form social communities. Existing social network models have been adopted by SIoT paradigm. The wide distribution of IoT objects and openness of social networks, however, make it more challenging to preserve privacy of IoT users. In this paper, we present a novel framework that preserves privacy against inference attacks on social network data through ranked retrieval models. We propose PVS, a privacy-preserving framework that involves the design of polymorphic value sets and ranking functions. PVS enables polymorphism of private attributes by allowing them to respond to different queries in different ways. We begin this work by identifying two classes of adversaries, authenticity-ignorant adversary, and authenticity-knowledgeable adversary, based on their knowledge of the distribution of private attributes. Next, we define the measurement functions of utility loss and propose PVSV and PVST that preserve privacy against authenticity-ignorant and authenticity-knowledgeable adversaries, respectively. We take into account the utility loss of query results in the design of PVSV and PVST. Finally, we show that PVSV and PVST meet the privacy guarantee with acceptable utility loss in extensive experiments over real-world datasets.

Reports on the 2013 AAAI Fall Symposium Series

AI Magazine ◽  
2014 ◽  
Vol 35 (2) ◽  
pp. 69-74
Author(s):  
Gully Burns ◽  
Yolanda Gil ◽  
Yan Liu ◽  
Natalia Villanueva-Rosales ◽  
Sebastian Risi ◽  
...  
Keyword(s):  
Artificial Intelligence ◽  
Neural Networks ◽  
Social Networks ◽  
Big Data ◽  
Computational Social Science ◽  
Social Contagion ◽  
Washington Dc ◽  
Web Analytics ◽  
Symbolic Representations ◽  
Symposium Series

The Association for the Advancement of Artificial Intelligence was pleased to present the 2013 Fall Symposium Series, held Friday through Sunday, November 15–17, at the Westin Arlington Gateway in Arlington, Virginia near Washington DC USA. The titles of the five symposia were as follows: Discovery Informatics: AI Takes a Science-Centered View on Big Data (FS-13-01); How Should Intelligence be Abstracted in AI Research: MDPs, Symbolic Representations, Artificial Neural Networks, or — ? (FS-13-02); Integrated Cognition (FS-13-03); Semantics for Big Data (FS-13-04); and Social Networks and Social Contagion: Web Analytics and Computational Social Science (FS-13-05). The highlights of each symposium are presented in this report.

scholarly journals Use of neural networks for forecasting of the exposure of social network users to destructive impacts

2020 ◽  
pp. 24-33
Author(s):  
Elena Doynikova ◽  
Aleksandr Branitskiy ◽  
Igor Kotenko
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Social Networks ◽  
Artificial Neural Networks ◽  
Social Network ◽  
Negative Impact ◽  
Young Generation ◽  
The Neural Network ◽  
Internet Environment ◽  
Artificial Neural

Introduction: In social networks, the users can remotely communicate, express themselves, and search for people with similarinterests. At the same time, social networks as a source of information can have a negative impact on the behavior and thinking oftheir users. Purpose: Developing a technique of forecasting the exposure of social network users to destructive influences, based onthe use of artificial neural networks. Results: A technique has been developed and experimentally evaluated for forecasting Ammon’stest results by a social network user’s profile using artificial neural networks. The technique is based on the results of Ammon’s testfor medical students. For training the neural network, a set of features was generated based on the information provided by socialnetwork users. The results of the experiments have confirmed the dependence between the data provided by social network users andtheir psychological characteristics. A mechanism has been developed aimed at prompt detection of destructive impacts or social networkusers’ profiles indicating the susceptibility to such impacts, in order to facilitate the work of psychologists. The experiments haveshown that out of the four investigated types of neural networks, the highest accuracy is provided by a multilayer neural network. Inthe future, it is planned to expand the set of features in order to achieve a better accuracy. Practical relevance: The obtained results canbe used to develop systems for monitoring the Internet environment, detecting the impacts potentially dangerous for mental health ofthe young generation and the nation as a whole.

Emerging Technologies in a Modern Competitive Scenario

2021 ◽  
pp. 1-16
Author(s):  
George Leal Jamil ◽  
Alexis Rocha da Silva
Keyword(s):  
Machine Learning ◽  
Neural Networks ◽  
Statistical Analysis ◽  
Differential Privacy ◽  
Social Issues ◽  
Training Data ◽  
Sensitive Data ◽  
Machine Learning Model ◽  
Highly Sensitive ◽  
Inference Attacks

Users' personal, highly sensitive data such as photos and voice recordings are kept indefinitely by the companies that collect it. Users can neither delete nor restrict the purposes for which it is used. Learning how to machine learning that protects privacy, we can make a huge difference in solving many social issues like curing disease, etc. Deep neural networks are susceptible to various inference attacks as they remember information about their training data. In this chapter, the authors introduce differential privacy, which ensures that different kinds of statistical analysis don't compromise privacy and federated learning, training a machine learning model on a data to which we do not have access to.

Sign in / Sign up

Export Citation Format

Share Document