Abstract: Social media tools have changed our world due to the way they convey information between individuals; this has led to many social movements either starting on social media or being organised and managed through this medium. At times however, certain human-induced events can trigger Human Security Threats such as Personal Security, Health Security, Economic Security or Political Security. The aim of this paper is to propose a holistic Data Analysis Framework for examining Social Movements and detecting pernicious threats to National Security interests. As a result of this, the proposed framework focuses on three main stages of an event (Detonating Event, Warning Period and Crisis Interpretation) to provide timely additional insights, enabling policy makers, first responders, and authorities to determine the best course of action. The paper also outlines the possible computational techniques utilised to achieve in depth analysis at each stage. The robustness and effectiveness of the framework are demonstrated by dissecting Warning Period scenarios, from real-world events, where the increase of Human Security aspects were key to identifying likely threats to National Security.

Abstract: Retweet prediction is a fundamental and crucial task in social networking websites as it may influence the process of information diffusion. Existing prediction approaches consider social network structure, but social context has not been fully considered. It is important to incorporate social contextual information into retweet prediction. Besides, the sparsity of retweet data severely disturb the performance of these models. In this paper, we propose a novel retweet prediction framework based on probabilistic matrix factorization model to integrate the observed retweet data, social influence and message embeddings semantic to improve the accuracy of prediction. We then incorporate these regularization terms into the objective function. Comprehensive experiments on the real-world dataset clearly validate both the effectiveness and efficiency of our model compared with several state-of the-art baselines.

Abstract: As one of the most important properties of the power grid, the voltage load plays an important role in the cascading failure of the smart grid and load redistribution can accelerate the speed of the failure by triggering more nodes to overload and fail. The subnet structure and different coupling modes also affect the robustness of the smart grid. However, the research on the effect of load, subnet structure and coupling mode on the cascading failure of the smart grid is still rare. In this paper, the smart grid with two-way coupling link consists of a power grid with small world topology and a communication network with scale- free topology. An improved load-capacity model is applied to overload- induced failure in the power grid and node importance (NI) is used as an evaluation index to assess the effect of nodes on the power grid and communication network. We propose three kinds of coupling modes based on NI of nodes between the cyber and physical subnets, i.e., Random Coupling in Subnets (RCIS), Assortative Coupling in Subnets (ACIS) and Disassortative Coupling in Subnets (DCIS). In order to improve the robustness of the smart grid, a cascading failure model based on load redistribution is proposed to analyze the influence of different coupling modes on the cascading failure of the smart grid under both a targeted attack and random attack. Some findings are summarized as: (I) The robustness of the smart grid is improved by increasing the tolerance α. (II) ACIS applied to the bottom-up coupling link is more beneficial in enhancing the robustness of the smart grid than DCIS and RCIS, regardless of a targeted attack or random attack.

Abstract: Volunteering and community reaction is known to be an essential part of response to critical events. With the rapid evolution of new means of communication, it has transformed accordingly. A new category of volunteers emerged – those that are not in the proximity to the area of emergency but willing to help the affected. Widely known as digital volunteers, they help aggregate, disseminate and distribute information to increase and maintain the awareness of stakeholders and resourceful individuals about the situation. There has been an upsurge of investigations of roles, timelines and aggregate characteristics of emergent communication. Compared to that, characteristics of crisis-related social media posts that predict wider social response to date have been studied modestly. In this research we are studying the process of reaction of potential digital volunteers to different extreme situations in social media platform.

Abstract: We present a novel computational pheromone model for describing dynamic network behaviors in terms of transition, persistency, and hosting. The model consists of a three-dimensional force-directed graph with bi-directional pheromone deposit and decay paths. A data compression algorithm is developed to optimize computational performance. We applied the model for visual analysis of a Malware Distribution Network (MDN), a connected set of maliciously compromised domains used to disseminate malicious software to victimize computers and users. The MDN graphs are extracted from crowdsourcing datasets from Google Safe Browsing (GSB) reports with attributions from VirusTotal report site. Our research shows that this novel approach reveals patterns of topological changes of the network over time, including the existence of persistent subnetworks and individual TLDs’ critical to the successful operation of MDNs, and the dynamics of the topological changes on daily basis. From the visualization, we observed notable clustering effects, and also noticed life span patterns for high edge count malware distribution clusters.