Vala Ali Rohani

Co-FAIS: Cooperative fuzzy artificial immune system for detecting intrusion in wireless sensor networks

Abstract Due to the distributed nature of Denial-of-Service attacks, it is tremendously challenging to identify such malicious behavior using traditional intrusion detection systems in Wireless Sensor Networks (WSNs). In the current paper, a bio-inspired method is introduced, namely the cooperative-based fuzzy artificial immune system (Co-FAIS). It is a modular-based defense strategy derived from the danger theory of the human immune system. The agents synchronize and work with one another to calculate the abnormality of sensor behavior in terms of context antigen value (CAV) or attackers and update the fuzzy activation threshold for security response. In such a multi-node circumstance, the sniffer module adapts to the sink node to audit data by analyzing the packet components and sending the log file to the next layer. The fuzzy misuse detector module (FMDM) integrates with a danger detector module to identify the sources of danger signals. The infected sources are transmitted to the fuzzy Q-learning vaccination modules (FQVM) in order for particular, required action to enhance system abilities. The Cooperative Decision Making Modules (Co-DMM) incorporates danger detector module with the fuzzy Q-learning vaccination module to produce optimum defense strategies. To evaluate the performance of the proposed model, the Low Energy Adaptive Clustering Hierarchy (LEACH) was simulated using a network simulator. The model was subsequently compared against other existing soft computing methods, such as fuzzy logic controller (FLC), artificial immune system (AIS), and fuzzy Q-learning (FQL), in terms of detection accuracy, counter-defense, network lifetime and energy consumption, to demonstrate its efficiency and viability. The proposed method improves detection accuracy and successful defense rate performance against attacks compared to conventional empirical methods.

Evaluation model for knowledge sharing in information security professional virtual community

Abstract Knowledge sharing has been proven to have affirmative effects on both the education and business sectors. Nevertheless, many professional virtual communities (PVC) have failed due to reasons, such as the low willingness of members to share knowledge with other members. In addition, it is not explicitly evident whether knowledge sharing in information security is able to reduce risk. To date, there have been relatively few empirical studies concerning the effects of knowledge sharing and its capability to reduce risk in information security communities. This paper proposes a model that is composed of two main parts. The first part is the Triandis theory, which is adapted to understand and foster the determinants of knowledge sharing behavior in PVCs. The second part explores the quantitative relationship between knowledge sharing and security risk reduction expectation. One hundred and forty-two members from the LinkedIn information security groups participated in this study. PLS analysis shows that perceived consequences, affect, and facilitating conditions have significant effects on knowledge sharing behavior. In contrast, social factors have shown insignificant effects on knowledge sharing behavior in information security communities. The results of the study demonstrate that there is a positive and strong relationship between knowledge sharing behavior and information security risk reduction expectation.

Sixty-four years of informetrics research: productivity, impact and collaboration

This paper analyses the information science research field of informetrics to identify publication strategies that have been important for its successful researchers. The study uses a micro-analysis of informetrics researchers from 5,417 informetrics papers published in 7 core informetrics journals during 1948–2012. The most productive informetrics researchers were analysed in terms of productivity, citation impact, and co-authorship. The 30 most productive informetrics researchers of all time span several generations and seem to be usually the primary authors of their research, highly collaborative, affiliated with one institution at a time, and often affiliated with a few core European centres. Their research usually has a high total citation impact but not the highest citation impact per paper. Perhaps surprisingly, the US does not seem to be good at producing highly productive researchers but is successful at producing high impact researchers. Although there are exceptions to all of the patterns found, researchers wishing to have the best chance of being part of the next generation of highly productive informetricians may wish to emulate some of these characteristics.

An Effective Recommender Algorithm for Cold-Start Problem in Academic Social Networks

Abundance of information in recent years has become a serious challenge for web users. Recommender systems (RSs) have been often utilized to alleviate this issue. RSs prune large information spaces to recommend the most relevant items to users by considering their preferences. Nonetheless, in situations where users or items have few opinions, the recommendations cannot be made properly. This notable shortcoming in practical RSs is called cold-start problem. In the present study, we propose a novel approach to address this problem by incorporating social networking features. Coined as enhanced content-based algorithm using social networking (ECSN), the proposed algorithm considers the submitted ratings of faculty mates and friends besides user’s own preferences. The effectiveness of ECSN algorithm was evaluated by implementing it in MyExpert, a newly designed academic social network (ASN) for academics in Malaysia. Real feedbacks from live interactions of MyExpert users with the recommended items are recorded for 12 consecutive weeks in which four different algorithms, namely, random, collaborative, content-based, and ECSN were applied every three weeks. The empirical results show significant performance of ECSN in mitigating the cold-start problem besides improving the prediction accuracy of recommendations when compared with other studied recommender algorithms.

Eliciting essential requirements for social networks in academic environments

Recent years have witnessed the generation of online social networking web sites, in which millions of members publicly articulate mutual friendship relations and share user-created contents. But it is just a few years that we have seen some efforts to create special social networks to be used in academic environments. Considering the increasing needs for this kind of web sites, we did a comprehensive survey on more than 20 academic social networks for gathering and categorizing the essential requirements for this kind of web sites. Hence, having some good experiences by creating the Iranian Experts Social Network with more than 120,000 official members, in this paper we try to clarify the features of Social Networks in academic environments and propose a category for their requirements.

Anomaly detection in Online Social Networks using structure-based technique

Online Social Networks as new phenomenon have affected our life in many positive ways; however it can be considered as way of malicious activities. Identifying anomalous users has become a challenge and many researches are conducted but they are not enough and in this paper we propose a methodology based on graph metrics of online social networks. The experimental results illustrate that majority of friends in online social networks have common friends with their friends while anomalous users may not follow this fact.

Utilizing machine learning in Sentiment Analysis: SentiRobo approach

Following the rapid evolution of Web 2.0, Sentiment Analysis has become one of the major techniques for mining the social media content. It aims to analyze opinions, sentiments, attitudes, and emotions towards entities such as topics, products, organizations, individuals, communities, and services. This paper presents SentiRobo, a supervised machine learning approach for the process of Sentiment Analysis. An enhanced version of Naive Bayes algorithm is introduced to predict the sentiment polarity of social media large data sets. Empirical evaluation over different twitter datasets with more than 300,000 records reveals the merit of this approach in processing of social media datasets.

An Enhanced Content-Based Recommender System for Academic Social Networks

The present study utilizes social computing techniques to enhance the content-based recommender systems. Coined as Enhanced Content-based Algorithm using Social Networking (ECSN), this recommender algorithm is applied in academic social networks to suggest the most relevant items to members of these online societies. In addition to considering users own preferences, ECSN takes advantage of the interest and preferences of users friends and faculty mates for providing more accurate recommendations. The research experiments were conducted by applying four different algorithms - random, collaborative, content-based, and ECSN, for 14 consecutive weeks. During this period, 1398 academic items were recommended to all 920 members of Malaysian Experts Academic Social Network (MyExpert). ANOVA tests indicate that the proposed algorithm significantly improves the prediction accuracy of algorithms based on well-known measurements of precision, fallout and F1. It is believed that this study can make a significant contribution to the level of user satisfaction in academic social networks.

Optimized-Memory Map-Reduce Algorithm for Mobile Learning

The increasing accessibility of mobile technologies and devices, such as smartphones and tablet PCs, has made mobile learning (m-learning) a critical feature of modern didactics. Mobile learning is among the many computerized activities that can be performed using mobile devices. As the volume of accessible important information on university websites continues to increase, students may face difficulties in accessing important information from a large dataset. This study introduces an algorithmic framework for data reduction that is built on optimized-memory map–reduce algorithm for mobile learning. The goal of this method is to generate meaningful recommendations to a collection of students in the easiest and fastest way by using a recommender system. Through an experiment, the proposed method has demonstrated significant improvements in data size reduction up to 77 %. Such improvements are greater than those that are achieved using alternate methods.