The use of online social networks is one of daily activity for people. OSNs' users share contents which sometimes has other users' information on. This type of data sharing is called co-owned data sharing and it is one of the most popular reasons for privacy leakages in OSNs platforms. OSNs' users claim that the responsibility of preserving users' privacy should be taken by those platforms. To do so, there are different approaches taken by OSNs for protecting users' privacy and providing the most secure environment to their users. This study proposes a model which should be adopted by OSNs platforms for a more secure environment. The model is developed on top of a hypothesis, the proof of the hypothesis is done with analysis on two questionnaires. Results of the analysis show that group decision making and reputation systems are needed to have more secure online social network platforms.
Group decision making (GDM) techniques have been very popular to take the best and the most convenient decision from an alternative set. GDM techniques have been applied in various area of information sharing as well as forensic information sharing. In a forensic investigation process, different forensic reports are produced by different investigators. Produced forensic reports are taken to the court as a file of evidence however deciding which report should be taken to the court is a challenging. Because decision is single handed, officer makes decision. There are two main approaches s/he might follow in selection; the first one is to combine the forensic reports into one report if investigators are located in the same environment, the other one is to choose the most comprehensive report (individually decided). Both approaches may cause continuance by undue process because of insufficient evidence because reports which are not taken to the court might include the needed evidence. In order to solve such problems in forensic area, this work provides a consensus-reached GDM approach in which extended induced ordered weighted average technique is used. Three forensic master students' forensic reports are used for application of the work. The result of the application phase showed that consensus-reached GDM makes the taken decision better and more accurate.
Vehicle-to-Grid (V2G) system is becoming a very popular concept since it has various benefits such as reducing energy consumption, being environmental friendly, bi-directional charging, and load balancing. Although, it gets highly remarkable and has many advantages, V2G system’s security is extremely challenging. Any security flaw in V2G system can cause serious issues on the system. Security issues might open doors to severe damages on the system. One of the most danger damage on such systems is disclosed confidential information. This work therefore analyses what are confidential information features in a V2G system, it then analyses whether a V2G system is vulnerable to attacks or not if the system’s confidential information is revealed. To do that, this study used fuzzy-classification technique in which a fuzzy system is developed. It also applied SVM and NB classification techniques in order to compare applied classification techniques in terms of their performances. Comparison results showed that fuzzy-classification technique performed better than other two techniques.
