Distributed Denial of Service (DDoS) attacks exhaust the resources of network services by generating a huge volume of network traffic. They constitute a primary threat to the current Internet community. To mitigate this threat, we propose a Machine Learning model based on Random Forest for detecting DDoS attacks. In our Random Forest, a great number of decision trees under both the Gini Index and Entropy criteria are constructed to improve the detection accuracy. Moreover, with the intrinsic simplicity of Random Forest model, our model is fast in terms of model convergence and attack detection. The data for building our model comes from the newly released dataset CICDDoS2019, which contains a large variety of DDoS attacks with a new classification based on network flows. The experimental results show that our model achieves high accuracy rates and F1_scores, and outperforms an existing state-of-the-art machine learning model for DDoS detection.

Interpretability in machine learning projects and one of its aspects - causal inference - have recently gained significant interest and focus. Due to the recent rapid appearance of frameworks, methods, algorithms and software most of which are in early stages of their development, it can be confusing for practitioners and researchers involved in a machine learning project to choose the best approach and set of techniques that would efficiently deliver valid insights while minimising the known risks of failure of data-related projects. CRISP-ML process methodology minimises this confusion by outlining a clear step-by-step process that explicitly treats of interpretability issues through every stage. The paper presents an update of CRISP-ML, which incorporates causality in a similar way and supports formalisation, design and implementation of specific instances of CRISP-ML process, subject to required levels of interpretability and causality of results. The approach is demonstrated on examples from the domains of credit risk, public health and healthcare.

Scatter-plot visualisation techniques are useful for analysing the correlations of variables on the axes as well as revealing patterns or abnormality in multidimensional data sets. However, Scatter-plot techniques have a drawback that they are not effective in showing a high number of dimensions where each plot in two-dimensional space can show a pair-wise of two variables on the X- and Y-axis. Star-plots are suitable for showing small data set with a low number of dimensions thank to the compactness in the visualisation. This paper proposes a hybrid technique that integrates Start-plots with Scatter-plots in the visualisation to enable the greater capability of scatter-plots in showing more information on each individual Star-plot. Our visualisation provides both overall views of mapping variables on multiple Scatter-plots whilst we also utilise Star-plots for showing the selected attributes on individual items for better comparison among and within variables. We also demonstrate the effectiveness of this hybrid method through case studies on various data sets.

Scatterplots and scatterplot matrix methods have been popularly used for showing statistical graphics and for exposing patterns in multivariate data. A recent technique, called Linkable Scatterplots, provides an interesting idea for interactive visual exploration which provides a set of necessary plot panels on demand together with interaction, linking and brushing. This article presents a controlled study with a mixed-model design to evaluate the effectiveness and user experience on the visual exploration when using a Sequential-Scatterplots who a single plot is shown at a time, Multiple-Scatterplots who number of plots can be specified and shown, and Simultaneous-Scatterplots who all plots are shown as a scatterplot matrix. Results from the study demonstrated higher accuracy using the Multiple-Scatterplots visualization, particularly in comparison with the Simultaneous-Scatterplots.​ While the time taken to complete tasks was longer in the Multiple-Scatterplots technique, compared with the simpler Sequential-Scatterplots, Multiple-Scatterplots is inherently more accurate. Moreover, the Multiple-Scatterplots technique is the most highly preferred and positively experienced technique in this study. Overall, results support the strength of Multiple-Scatterplots and highlight its potential as an effective data visualization technique for exploring multivariate data.

Genomic data are large and complex which are challenges to visualize them effectively on ordinary screens due to the limited display spaces. Large and high resolution displays could enable the capability to show more information at once for better comprehension from the visualization. This paper presents a two-dimensional interactive visualization system and supporting algorithm for multi-dimensional large genomic data analysis that can be used in both ordinary displays or immersive environments. We provide both view of the entire patient cohort in the similarity space and the genomic details currently for comparison among the patients. Through the similarity space and on the selected genes of interest, we are able to perceive the genetic similarity throughout the cohort. From the linked heat map visualisation of the selected genes, we apply hierarchical clustering on both the horizontal and vertical axes to group together the genetically similar patients. We demonstrate the effectiveness of the visualization with two case studies on pediatric cancer patients suffering from Acute Lymphoblastic Leukemia (ALL) and from Rhabdomyosarcoma (RMS).

Graphic game engines have introduced even more advanced technologies to improve the rendering, image quality, ergonomics, and user experience of their creations by providing user-friendly yet powerful tools to design and develop new games. There are thousands of genes in the human genome that contain information about specific individual patients and the biological mechanisms of their diseases. The complexity in biomedical and genomic data usually requires effective visual information processing and analytics. Unfortunately, available visualisation techniques for this domain are limited, many in static forms. The open study questions here are as follow: Are there lessons to be learnt from these video games? Or could the game technology help us explore new graphic ideas accessible to non-specialists? This paper presents a visual analytics model that enables the analysis of large and complex genomic data using Unity3D game technology. This includes an interactive visualisation, providing an overview of the patient cohort with a detailed view of the individual genes. We illustrate the effectiveness of our approach in guiding the effective treatment decision in the cohort through datasets from the childhood cancer B-Cell acute lymphoblastic leukaemia.

Speech and language researchers often rely on large, naturalistic, audio-visual corpora to identify and measure patterns of language structure, variation, change and use. There are, however, few visualization tools designed for this need. This paper proposes a novel visual analytic method to process large linguistic corpora by employing Bayes' Theorem and interactive visualization. We adopt a simple and meaningful design in our visualization for linguists to understand. Instead of offering a fixed visualization, this project enables greater interaction through filtering, grouping and dragging. Multiple phases are included in the system, from processing the metadata exported from popular standalone linguistic software, to creating the visualization, and enabling interaction and filtering.

Science fiction media has had a long lasting influence on the progression of interactive technology, however recently contradictions are emerging in the development of the two disciplines. Therefore, in this exploratory position paper we report on the insights attained through a day long workshop amongst scientists and researchers on how the collaboration between science fiction and Human Computer Interaction (HCI) can be advanced. Discussions in the workshop focused on detailing the relationship between HCI and science fiction. In conclusion, as our main contribution an action plan and agenda is presented for facilitating deeper influences amongst the two disciplines.

Clarity, simplicity and visual adjustability to the preference of the analyst are key aspects of the visualization techniques required by visual analytics in broad sense. Scatterplots and scatterplot matrices are commonly used for visually analyzing multidimensional multivariate data. This paper presents a new approach for deep visual exploration of large multi-attribute data using linkable scatterplots. Proposed method overcomes the limitations of the single scatterplot by providing more plot panels for better comparison while it reduces the unnecessary number of panels of the scatterplot matrix method. The panels are fully interactive and linking together where variables can be mapped on axes independently or on common visual attributes such as color, size and shape. We illustrate the effectiveness of proposed linkable scatterplot method on various data sets.

In this paper we investigate the concept of believability and make an attempt to isolate individual characteristics (features) that contribute to making virtual characters believable. As the result of this investigation we have produced a formalisation of believability and based on this formalisation built a computational framework focused on simulation of believable virtual agents that possess the identified features. In order to test whether the identified features are, in fact, responsible for agents being perceived as more believable, we have conducted a user study. In this study we tested user reactions towards the virtual characters that were created for a simulation of aboriginal inhabitants of a particular area of Sydney, Australia in 1770 A.D. The participants of our user study were exposed to short simulated scenes, in which virtual agents performed some behaviour in two different ways (while possessing a certain aspect of believability vs. not possessing it). The results of the study indicate that virtual agents that appear resource bounded, are aware of their environment, own interaction capabilities and their state in the world, agents that can adapt to changes in the environment and exist in correct social context are those that are being perceived as more believable. Further in the paper we discuss these and other believability features and provide a quantitative analysis of the level of contribution for each such feature to the overall perceived believability of a virtual agent.

A large number of tracking and gesture recognition algorithms and technologies have been developed in the field of human-computer interactions thanks to the introduction of cameras with depth sensors such as Microsoft's Kinect. Most of the techniques rely on skeleton tracking which is more suitable for distant and full body interaction. This paper presents a new real-time finger-gesture interaction system using Kinect v2 that identifies fingertips and finger gestures that enable the natural user interaction at a close distance. Our contribution also includes various gesture recognition algorithms using two and three fingers such as L-gesture, OK-gesture, Rock-gesture and Scissor-gesture, in addition to full hand and one-finger gestures. We demonstrate the effectiveness of our system through a fruit slicing game.

Interactive visual analysis can support displaying complex multidimensional data. We developed a novel system, which can incorporate domain knowledge and visual guidelines, to present complex logistical data. This paper presents the details of the development process of that system. Diagrammatic visualization approach was adopted for presenting multidimensional port data, in relation to events. The paper discusses case studies with land-side data and wharf-side data obtained from Port Botany in Sydney, Australia. We conducted a usability study with 20 students that compared performance with our visualization against a traditional bar chart. Task completion was significantly shorter with the diagrammatic visualization, and the visualization was rated higher in overall user preference.

This paper extends a previous work on node link tree visualization and interaction by providing visual clues on hidden structures. We adopt the effectiveness of DOI Tree, a multi-focal tree layout algorithm, for exploring large hierarchical structures. The advantages of visualization are its most familiar mapping for users, its capability on providing multiple focused nodes, and its dynamic rescaling of substructures to fit the available space. By providing various methods of topological previews of substructure including simple icon view, tree view and tree map view, we provide better understanding the topology of hidden branches.

Meetings are an integral part of the work place and society in general. Research in Computer Supported Cooperative Work attempts to facilitate and make the process of meetings more effective. Our vision is that the incorporation of social robots in such human-human collaborative settings can assist and improve the effectiveness of a meeting. In this paper we present an empirical study in which pairs of participants collaborate in a meeting scenario with a Nao humanoid robot. Using a within-subjects design, we manipulated the robot's role within the meeting as being either “active” versus “passive”/“service-oriented”. Our results show that the more active robot was deemed as more more alive and social, had the participants more emotionally involved and caused more verbal engagement from the participants as compared to a passive service robot. In conclusion, we speculate on the inclusion of a collaborative robot as a meeting partner.

It is becoming increasingly difficult to stay aware of the state-of-the-art in any research field due to the exponential increase in the number of academic publications. This problem effects authors and reviewers of submissions to academic journals and conferences, who must be able to identify which portions of an article are novel and which are not. Therefore, having a process to automatically judge the flow of novelty though a document would assist academics in their quest for truth. In this article, we propose the concept of Within Document Novelty Location, a method of identifying locations of novelty and non-novelty within a given document. In this preliminary investigation, we examine if a second order statistical model has any benefit, in terms of accuracy and confidence, over a simpler first order model. Experiments on 928 text sequences taken from three academic articles showed that the second order model provided a significant increase in novelty location accuracy for two of the three documents. There was no significant difference in accuracy for the remaining document, which is likely to be due to the absence of context analysis.

While prior works on enclosure approach, guarantees the space utilization of a single geometrical area, mostly rectangle, this paper proposes a flexible enclosure tree layout method for partitioning various polygonal shapes that break through the limitation of rectangular constraint. Similar to Treemap techniques, it uses enclosure to divide display space into smaller areas for its sub-hierarchies. The algorithm can partition a polygonal shape or even an arbitrary shape into smaller polygons, rotated rectangles or vertical-horizontal rectangles. The proposed method and implementation algorithms provide an effective interactive visualization tool for partitioning large hierarchical structures within a confined display area with different shapes for real-time applications. We demonstrated the effective of the new method with a case study, an automated evaluation and a usability study.

In this paper we present preliminary pilot study of how people's interactions can be characterized in open-ended environments through the concept of engagement. By open - ended environments we refer to physical spaces that construct content and associated semantics through non-didactic methods supportive of non - linear navigation. This work presents an overview of an approach to interactions as mapping human actions and their characteristics, through which behavior features can be identified. The interpretation is based on the extension of previous work, the Kinetic Inter-Acting System, that interprets, visualises and offers means for analysis of the interaction process between parties enabling visual reasoning about the quality of interactions. The approach is examined in the context of new developments in museums that perceives them as becoming creative and reflective agents in digital humanities. Examples are taken from a large public immersive exhibition that relies upon interaction delivered through various modalities for content assimilation and participant experience.

Space-filling visualization techniques have proved their capability in visualizing large hierarchical structured data. However, most existing techniques restrict their partitioning process in vertical and horizontal direction only, which cause problem with identifying hierarchical structures. This paper presents a new space-filling method named Angular Treemaps that relax the constraint of the rectangular subdivision. The approach of Angular Treemaps utilizes divide and conquer paradigm to visualize and emphasize large hierarchical structures within a compact and limited display area with better interpretability. Angular Treemaps generate various layouts to highlight hierarchical sub-structure based on user's preferences or system recommendations. It offers flexibility to be adopted into a wider range of applications, regarding different enclosing shapes. Preliminary usability results suggest user's performance by using this technique is improved in locating and identifying categorized analysis tasks.

Simulating large crowds of virtual agents has become an important problem in virtual reality applications, video games, cinematography and training simulators. In this paper, we show how to achieve a high degree of appearance variation among individual 3D avatars in generated crowds through the use of genetic algorithms, while also manifesting unique characteristic features of a given population group. We show how virtual cities can be populated with diverse crowds of virtual agents that preserve their ethnic features, illustrate how our approach can be used to simulate full body avatar appearance, present a case study and analyze our results.

Interactions are central to our endeavours. Whether in a face-to-face session with a doctor or via one of Web 2.0 applications, the way interactions unfold has an impact on the final outcome of the session. This paper addresses the problem of the development of a visual language and analysis system, enabling visual reasoning about the quality of interactions. The paper presents a system of encoding the way interactions unfold. It presents the foundations of KIA (Kinetic Inter-Acting) visual language and the visual analysis system with this language. The language is then evaluated through comparison with expert judgments when applied to cases in health care. KIA (the language and the analysis system) enables both humans and machines to utilise information about how interactions unfold in order to improve interactions and the processes that depend on them.

With an astonishing amount of data generated for processing on a daily basic, it is essential to provide an effective methodology for understanding, reasoning and supporting decision making of large information spaces. This paper presents a new concept that provides an intelligent and interactive visualization in supporting large scale analysis. This aims to provide a much greater flexibility and control for the users to interactively customize the visualizations according to their preferences. A simple prototype is also presented to demonstrate the concept on hierarchical structures.

Contemporary Web-based electronic markets reflect the dominating content-based systems approach of Web 2.0. Though useful, these electronic markets are far from being believable trading places. Marketplace is where things and traders have presence, constituting a rich interaction space. The believability of the place depends on the believability of the presence and interactions in it, including the players' behaviour and the narrative scenarios of the marketplace. This paper discusses what constitutes the believability of electronic marketplaces and presents the technologies that support it. Believability of electronic marketplaces can be described through three metaphors: ``marketplaces where people are'', ``marketplaces that are alive and engaging'', and ``market places where information is valuable and useful''. The paper presents the core technologies that enable the perceivable believability of electronic marketplaces. It describes a demonstrable prototype of a Web-based electronic marketplace that integrates these technologies. This is part of a larger project that aims to make informed automated trading an enjoyable reality of Web 3.0.

Our research combines electronic institutions and 3D virtual worlds for the construction of virtual institutions which are virtual worlds with normative regulation of interactions. That is, a virtual world where participants actions have to comply with predefined institutional rules. In this context, the actions a participant may perform depend on the institutional rules and the current execution state. We propose to include iObjects, intelligent objects, as entities having both visualization properties and decision mechanisms in the virtual institution. They are a new key element to improve users participation in virtual institutions. We situate them in a middleware infrastructure in order to be independent of 3D virtual world platform and to provide a general solution in which participants could be connected from different immersive environment platforms.

First, a hierarchical modelling method, VQSVM, is introduced, and some remarks are discussed. Secondly the proposed VQSVM is applied to a nonstandard learning environment, imbalanced data sets. In cases of extremely imbalanced dataset with high dimensions, standard machine learning techniques tend to be overwhelmed by the large classes. The hierarchical VQSVM contains a set of local models i.e. codevectors produced by the vector quantization and a global model, i.e. support vector machine, to rebalance datasets without significant information loss. Some issues, e.g. distortion and support vectors, have been discussed to address the trade-off between the information loss and undersampling rate. Experiments compare VQSVM with random resampling techniques on some imbalanced datasets with varied imbalance ratios, and results show that the performance of VQSVM is superior or equivalent to random resampling techniques, especially in case of extremely imbalanced large datasets.

The paper discusses an approach of using traditional time series analysis, as domain knowledge, to help the data-preparation of support vector machine for classifying documents. Classifying unexpected news impacts to the stock prices is selected as a case study. As a result, we present a novel approach for providing approximate answers to classifying news events into simple three categories. The process of constructing training datasets is emphasized, and some time series analysis techniques are utilized to pre-process the dataset. A rule-base associated with the net-of-market return and piecewise linear fitting constructs the training data set. A classifier mainly built by support vector machine uses the training data set to extract the interrelationship between unexpected news events and the stock price movements.

Consumer reviews, opinions and shared experiences in the use of a product is a powerful source of information about consumer preferences that can be used in recommender systems. Despite the importance and value of such information, there is no comprehensive mechanism that formalizes the opinions selection and retrieval process and the utilization of retrieved opinions due to the difficulty of extracting information from text data. In this paper, a new recommender system that is built on consumer product reviews is proposed. A prioritizing mechanism is developed for the system. The proposed approach is illustrated using the case study of a recommender system for digital cameras

Fully automated trading, such as e-procurement, using the Internet is virtually unheard of today. Three core technologies are needed to fully automate the trading process: data mining, intelligent trading agents and virtual institutions in which informed trading agents can trade securely both with each other and with human agents in a natural way. This paper describes a demonstrable prototype e-trading system that integrates these three technologies and is available on the World Wide Web. This is part of a larger project that aims to make informed automated trading a reality.

Collaborative virtual environments are becoming an intrinsic part of professional practices. In addition to providing collaboration support, they have the potential to collect vast amounts of data about collaborative activities. The aim of this research is to utilize this data effectively, extract meaningful insights out of it and feeding discovered knowledge back into the environment. The paper presents a framework for integrating knowledge discovery techniques with collaborative virtual environments, starting from early conceptual development. Discovered patterns are deposited in an organizational memory which makes these available within the virtual environment. Two examples of the application of the framework are included.

No abstract available.