No abstract

The purpose of this study is to investigate the effects of contextual features on automatic detection accuracy of online recruitment frauds in Australian job market. In addition, the study aims to unearth the significance of localisation of such approaches. The study first generates a dataset based on a local and semi-structured advertising platform in Australia. The labelled dataset is then used to train a learning model on several content-based and contextual features. The existence of advertising body in relevant government and non-government registries in Australia, along with the internet presence of the advertiser, were considered as contextual features. The extraction process of such contextual features was automated as well. The study concludes that the inclusion of contextual features improves the performance measures of the automated online recruitment fraud detection model. The practical implication of the study is two-folds. Firstly, the contextual feature space generation engine can be used with any dataset, with minimal localisation efforts. Secondly, such learning models can be used in the back end of online job recruitment portals to detect and prevent online recruitment frauds. The study not only demonstrates the positive impact of using contextual features in fraud detection using a real-life dataset, but it also demonstrates how these contextual features can be extracted automatically from the web, based on localised company registries. The generated dataset has been made available for further research in the domain.

The purpose of this paper is to analyse the effects of predatory approach words in the detection of cyberbullying and to propose a mechanism of generating a dictionary of such approach words. The research incorporates analysis of chat logs from convicted felons, to generate a dictionary of sexual approach words. By analysing data across multiple social networks, the study demonstrates the usefulness of such a dictionary of approach words in detection of online predatory behaviour through machine learning algorithms. It also shows the difference between the nature of contents across specific social network platforms. The proposed solution to detect cyberbullying and the domain of approach words are scalable to fit real-life social media, which can have a positive impact on the overall health of online social networks. Different types of cyberbullying have different characteristics. However, existing cyberbullying detection works are not targeted towards any of these specific types. This research is tailored to focus on sexual harassment type of cyberbullying and proposes a novel dictionary of approach words. Since cyberbullying is a growing threat to the mental health and intellectual development of adolescents in the society, models targeted towards the detection of specific type of online bullying or predation should be encouraged among social network researchers.

Due to the proliferation of data and services in everyday life, we face challenges to ascertain all the necessary contexts and associated contextual conditions and enable applications to utilize relevant information about the contexts. The ability to control context-sensitive access to data resources has become ever more important as the form of the data varies and evolves rapidly, particularly with the development of smart Internet of Things (IoTs). This frequently results in dynamically evolving contexts. An effective way of addressing these issues is to model the dynamically changing nature of the contextual conditions and the transitions between these different dynamically evolving contexts. These contexts can be considered as different states and the transitions represented as state transitions. In this paper, we present a new framework for context-sensitive access control, to represent the dynamic changes to the contexts in real time. We introduce a state transition mechanism to model context changes that lead the transitions from initial states to target states. The mechanism is used to decide whether an access control decision is granted or denied according to the associated contextual conditions and controls data access accordingly. We introduce a Petri net model to specify the control flows for the transitions of states according to the contextual changes. A software prototype has been implemented employing our Petri net model for detection of such changes and making access control decisions accordingly. The advantages of our context-sensitive access control framework along with a Petri net model have been evaluated through two sets of experiments, especially by looking for re-evaluation of access control decisions when context changes. The experimental results show that having a state transition mechanism alongside the context-sensitive access control increases the efficiency of decision making capabilities compared to earlier approaches.

During the onset of COVID-19 pandemic, the social media was flooded with misinformation. Irrespective of the type of the misinformation, such contents played a significant role in increasing confusion among people in the middle of an ongoing crisis. The purpose of the study is to investigate the nature of a specific type of misinformation, i.e., rumors, surrounding COVID-19. The study utilizes a publicly available and labelled Twitter dataset and proposes a novel feature space, which can detect rumor instances with high accuracy. The proposed feature space not only includes content-based features, but also includes psycholinguistic features to further study the characteristics of the content from the perspectives of linguistics and psychology. The use of psycho-linguistic features has been utilised to understand certain dramatisation of text in the domain of conspiracy propagation and fake news detection. However, the use of such dramatisation detection approach has never been used for the purposes of rumor detection. Our study first outlines the differences between these categories of misinformation propagation and clarifies where rumor fits-in under the broader umbrella of misinformation. It further outlines how the use of psycho-linguistic features can also improve the detection accuracy of rumors on social media. The study demonstrates through multiple experimental setups that psycho-linguistic features improves the detection accuracy and associated performance measures, such as precision and recall, for COVID-19 rumors on Twitter. The observed improvements are consistent across multiple machine learning models.