Soran Ibrahim; Qing Tan
Abstract
In the recent years, social networks (SN) are now employed for communication and networking, socializing, marketing, as well as one’s daily life. Billions of people in the world are connected though various SN platforms and applications, which results in generating massive amount ...
Read More
In the recent years, social networks (SN) are now employed for communication and networking, socializing, marketing, as well as one’s daily life. Billions of people in the world are connected though various SN platforms and applications, which results in generating massive amount of data online. This includes personal data or Personally Identifiable Information (PII). While more and more data are collected about users by different organizations and companies, privacy concerns on the SNs have become more and more prominent. In this paper, we present a study on information privacy in SNs through exploring the general laws and regulations on collecting, using and disclosure of information from Canadian perspectives based on the Personal Information Protection and Electronic Document Act (PIPEDA). The main focus of this paper is to present results from a survey and the findings of the survey.
Mohammad Reza Mohammadrezaei; Mohammad Ebrahim Shiri; Amir Masoud Rahmani
Abstract
Detection of fake accounts on social networks is a challenging process. The previous methods in identification of fake accounts have not considered the strength of the users’ communications, hence reducing their efficiency. In this work, we are going to present a detection method based on the users’ ...
Read More
Detection of fake accounts on social networks is a challenging process. The previous methods in identification of fake accounts have not considered the strength of the users’ communications, hence reducing their efficiency. In this work, we are going to present a detection method based on the users’ similarities considering the network communications of the users. In the first step, similarity measures somethings such as common neighbors, common neighbors graph edges, cosine, and the Jaccard similarity coefficient are calculated based on adjacency matrix of the corresponding graph of the social network. In the next step, in order to reduce the complexity of data, Principal Component Analysis is applied to each computed similarity matrix to provide a set of informative features. then, a set of highly informative eigenvectors are selected using elbow-method. Extracted features are employed to train a One Class Classification (OCC) algorithm. Finally, this trained model is employed to identify fake accounts. As our experimental results indicate the promising performance of the proposed method a detection accuracy and false negative rates are 99.6% and 0%, respectively. We conclude that bringing similarity measures and One Class Classification algorithms into play, rather than the multi-class algorithms, provide better results.
H. Mohammadhassanzadeh; H. R. Shahriari
Abstract
In Social networks, users need a proper estimation of trust in others to be able to initialize reliable relationships. Some trust evaluation mechanisms have been offered, which use direct ratings to calculate or propagate trust values. However, in some web-based social networks where users only have ...
Read More
In Social networks, users need a proper estimation of trust in others to be able to initialize reliable relationships. Some trust evaluation mechanisms have been offered, which use direct ratings to calculate or propagate trust values. However, in some web-based social networks where users only have binary relationships, there is no direct rating available. Therefore, a new method is required to infer trust values in these networks. To bridge this gap, this paper aims to propose a new method which takes advantage of user similarity to predict trust values without any need for direct ratings. In this approach, which is based on socio-psychological studies, user similarity is calculated from the profile information and the texts shared by the users via text-mining techniques. Applying Ziegler ratios to our approach revealed that users are more than 50% more similar to their trusted agents than to arbitrary peers, which proves the validity of the original idea of the study about inferring trust from language similarity. In addition, comparing the real assigned ratings, gathered directly from users, with the experimental results indicated that the predicted trust values are sufficiently acceptable (with a precision of 61%). We have also studied the benefits of using context in inferring trust. In this regard, the analysis revealed that the precision of the predictions can be improved up to 72%. Besides the application of this approach in web-based social networks, the proposed technique can also be of much help in any direct rating mechanism to evaluate the correctness of trust values assigned by users, and increase the robustness of trust and reputation mechanisms against possible security threats.