As our dependence on intelligent machines continues to grow, so does the demand for more transparent and interpretable models. In addition, the ability to explain the model generally is now the gold standard for building trust and deployment of artificial intelligence systems in critical domains. Explainable artificial intelligence (XAI) aims to provide a suite of machine learning techniques that enable human users to understand, appropriately trust, and produce more explainable models. Selecting an appropriate approach for building an XAI-enabled application requires a clear understanding of the core ideas within XAI and the associated programming frameworks. We survey state-of-the-art programming techniques for XAI and present the different phases of XAI in a typical machine learning development process. We classify the various XAI approaches and, using this taxonomy, discuss the key differences among the existing XAI techniques. Furthermore, concrete examples are used to describe these techniques that are mapped to programming frameworks and software toolkits. It is the intention that this survey will help stakeholders in selecting the appropriate approaches, programming frameworks, and software toolkits by comparing them through the lens of the presented taxonomy.

Industrial cyber-physical systems ( ICPSs ) manage critical infrastructures by controlling the processes based on the “physics” data gathered by edge sensor networks. Recent innovations in ubiquitous computing and communication technologies have prompted the rapid integration of highly interconnected systems to ICPSs. Hence, the “security by obscurity” principle provided by air-gapping is no longer followed. As the interconnectivity in ICPSs increases, so does the attack surface. Industrial vulnerability assessment reports have shown that a variety of new vulnerabilities have occurred due to this transition. Although there are existing surveys in this context, very little is mentioned regarding the outputs of these reports. While these reports show that the most exploited vulnerabilities occur due to weak boundary protection, these vulnerabilities also occur due to limited or ill-defined security policies. However, current literature focuses on intrusion detection systems ( IDSs ), network traffic analysis ( NTA ) methods, or anomaly detection techniques. Hence, finding a solution for the problems mentioned in these reports is relatively hard. We bridge this gap by defining and reviewing ICPSs from a cybersecurity perspective. In particular, multi-dimensional adaptive attack taxonomy is presented and utilized for evaluating real-life ICPS cyber incidents. Finally, we identify the general shortcomings and highlight the points that cause a gap in existing literature while defining future research directions.

Little is currently known about the factors that promote the propagation of information in online social networks following terrorist events. In this paper we took the case of the terrorist event in Woolwich, London in 2013 and built models to predict information flow size and survival using data derived from the popular social networking site Twitter. We define information flows as the propagation over time of information posted to Twitter via the action of retweeting. Following a comparison with different predictive methods, and due to the distribution exhibited by our dependent size measure, we used the zero-truncated negative binomial (ZTNB) regression method. To model survival, the Cox regression technique was used because it estimates proportional hazard rates for independent measures. Following a principal component analysis to reduce the dimensionality of the data, social, temporal and content factors of the tweet were used as predictors in both models. Given the likely emotive reaction caused by the event, we emphasize the influence of emotive content on propagation in the discussion section. From a sample of Twitter data collected following the event (N = 427,330) we report novel findings that identify that the sentiment expressed in the tweet is statistically significantly predictive of both size and survival of information flows of this nature. Furthermore, the number of offline press reports relating to the event published on the day the tweet was posted was a significant predictor of size, as was the tension expressed in a tweet in relation to survival. Furthermore, time lags between retweets and the co-occurrence of URLS and hashtags also emerged as significant.

Handoff mechanisms allow mobile users to move across multiple wireless access points while maintaining their voice and/or data sessions. A traditional handoff process is concerned with smoothly transferring a mobile device session from its current access point (or cell) to a target access point (or cell). These handoff characteristics are sufficient for voice calls and background data transfers, however nowadays many mobile applications are heavily based on data and processing capabilities from the cloud. Such applications, especially those that require greater interactivity, often demand not only a smooth session transfer, but also the maintenance of quality of service requirements that impact a user's experience. In this context, the Fog Computing paradigm arises to overcome delays encountered when applications need low latency to access data or offload processing to the cloud. Fog computing introduces a distributed cloud layer, composed of cloudlets (i.e., "small clouds" with lower computational capacity), between the user and the cloud. Cloudlets allow low latency access to data or processing capabilities, which can be accomplished by offering a VM to the user. An overview of Fog computing is first providing, relating it to general concepts in Cloud-based systems, followed by a general architecture to support virtual machine migration in this emerging paradigm - discussing both the benefits and challenges associated with such migration.

Deep Neural Networks (DNNs) have been successfully used in a number of application areas. As the number of layers and neurons in DNNs increases rapidly, significant computational resources are needed to execute a DNN. This ever-increasing resource demand is currently met by large-scale data centers with state-of-the-art GPUs. However, increasing availability of mobile edge computing and 5G technologies provide new possibilities for DNN-driven AI applications, especially where these application make use of data sets that are distributed in different locations. One fundamental process of a DNN-driven application in mobile edge clouds is the adoption of ``inferencing'' - the process of executing a pre-trained DNN based on newly generated data from mobile devices. We investigate offloading DNN inference requests in a 5G-enabled mobile edge cloud (MEC), to admit as many inference requests as possible. We propose exact and approximate solutions for the problem. We also consider dynamic task offloading for inference requests, and devise an online algorithm that can be adapted in real time. The proposed algorithms are evaluated through large-scale simulations and using a real world test-bed. The experimental results demonstrate that the empirical performance of the proposed algorithms outperform their theoretical counterparts and other similar heuristics reported in literature.